This action might not be possible to undo. Are you sure you want to continue?
_
COMPREHENIIVE
vvvvOfvvvv
MATHEMATICS
Chief Editor & Compiler:
Roger Thompson
ABHISHEK
All rights reserved. No part of this book may be reproduced in any
form, electronically or otherwise, in print, photoprint, micro film or
by any other means without written permission from the publi sher.
ISBN
Copyright
Revised Edition
Published by
Publisher
2010
ABHISHEK PUBLICATIONS,
S.C.O. 5759, Sector 17C,
CHANDIGARH1600 17 (India)
Ph.2707562,FaxOI722704668
Email: abhpub@yahoo.com
9788182473416
Preface
Mathematicians, scientists, engineers, and students will
look this dictionary for defmitive coverage of all branches
of mathematics, both pure and applied. Featuring more
than 1500 terms, the book defmes terms and expressions
in algebra, number theory, operator theory, logic, com
plex numbers, fmite mathematics, topology, and other ar
easeach with definition along with pictorial represen
tations of many terms.
This dictionary is authoritative, comprehensive compris
ing latest terms and is carefully reviewed to ensure its
accuracy, clarity, and completeness. This Dictionary of
Mathematics puts a wealth of essential information at
your fingertips. Whether you're a professional, a stu
dent, a writer, or a general reader with science curiosity,
this comprehensive resource defines the current language
of pure and applied Mathematics and gives you a better
understanding of the ideas and concepts you need to
know.
It has, for the first time, brought together in one easily
accessible form the bestexpressed thoughts that are es
pecially illuminating and pertinent to the discipline of
Mathematics. This dictionary will be a handy reference
for the mathematician or scientific reader and the wider
public interested in who has said what on mathematics.
· The overall aim of the dictionary is to provide an acces
sible description of what one judges to be the core ma
terial for damn good dictionary. The subject has a repu
tation for being disagreeably difficult. I have tried to
alter that perception by showing that key ideas can be
presented simply judiciously but not overwhelmingly de
ployed, clarifies and provides power.
11
2
by2 table I abeliangroup
A ~
T 45'
0.
.2by2 table .
1. this is a twoway table ;
where the numbers of levels .
of the ro·.v and columnclassi I
fications are each 2. If the row I
and column classifications
each divide the observational
units into subsets, then it is I
likely that it will be useful to
analyse the data using the
Fisher Test.
~ .8
5
• 3D figure
a set of points in space; ex
amples: box, cone, cylinder,
parallelpiped, prism, pyramid,
regular pyramid, right cone,
right cylinder, right prism,
sphere.
; 'prime'; designates an image
: corresponding to the preimage
~ using the same variable.
• 454590 triangle
an isoscoles right triangle
I
.• aRb
I a is an element in b
I. abacus
~ a Japanese counting device and
: calculator.
~ • abelian group
I a group in which the binary
f
°',;'" rIO;' ..
.'1... .._'1 ~ . i f . . . . . . •. .if. ...... ( ~ , '=)
Wi'Z" w;;g,
!!!!!6=" ========!!!!!!!!* sbseisss IIIJlinetnlnsj'tmnllritm II
is commutative, that • additive identity property
IS,. ab= ba for all elements a abd ; the sum .pf any number and
b m the group. : zero is the original number·
• abscissa zero is the identity element of
the xcoordinate of a point in a addition.
2dimensional coordinate sys ; • adjacent angles
tern ..gh
. ; two nonstraJ. t and nonzero
• absolute value : angles that have a common side
the positive value for a real in the interior of the angle
disregarding the sign. ; f?rmed by the nons:ommon
Wntten 1 x I. For example, : SIdes
131 =3, 141 =4, and 101 =0. lnIeriorangl .... nlhe
....... side oflhe InirIlIversaI A IntedonI on the ......
• abundant number ...
a positive integer that is smaller ;
than the sum of its proper divi
sors. .
Be 0 DE 1IIC3o_18D
• acceleration .
rate of change of velocity ; • interior angle
With respect to time. : the mtenor angle that forms a
• "' .... .....hl I els f linear pair with a given exterior
e eVi 0 accuracy I angle of a triangle.
determined by the : . ."
sItuation or the given numbers I· adjacent SIde
stude?ts should help (of an angle.in a one
what IS acceptable according to : of the two SIdes of the mangle
the situation. that form the sides of the angle.
• accuracy ; • affine superimposition
the closeness of a measurement ; a for which the
or estimate to its true value. : assocIated transformations are
I all affine.
• acute angle :
an angle whose measure is • affine transforma,tion
greater than 0 but less than 90 ; a transformation for which
degrees. : parallel lines remain parallel.·
I
llilleeimlie elJllllhon lilleeimlie numb:..
was applied to landmark
; based shape .
Mfine transformations of the
plane take squares into paral
lelograms and take circles into
ellipses of the same shape. M I 'g
fine transformations of a 3di I :or
mensional space take cubes
into parallelopipeds (sheared
bricks) and spheres into ellip I
soids all of the same shape.
Similar results are produced
in higher dimensional spaces. I
Equivalent to "uniform trans
formation" . As far as form is
concerned' (that is, ignoring
translation and rotation), any
affine transformation can be
'.
L ' . '
Y'II .1 "8; _:l ____
, , '
: .t,. ....
x,.
x"
 algebraic equation
I an equation of the form
f( x) = 0 where f is a polyno
: mial.
diagrammed as a pure strain I
taking a square to a rectangle
on the same axes. In studies
of shape, where scale is ig I
nored as well, the picture is the I
same but now the sum of the
squares of the axes is un
changing. Still ignoring scale I
(that is, as far as shape is con
cerned), any affine transfor
mation can be also dia I
grammed as a pure shear tak  I
tPq
where
Wq
q
tlq
and
6
=
Wq • 1l.y • tlq
=
q. eq
=
(ra  i) . ax + q.m
=
!6f. st
n
=
ing a square into a parallelo : _ algebraic number
gram of unchanged base seg a number that is the root of
ment and height. This dia ; an algebraic polynomial. For
gram of shear came into : example, (sqrt 2) is an alge
morphometrics via an applica braic number because it is a
tion to principal components solution of the equation x
2
=2.
analysis somewhat before it
II.
tdgorithm I RltemateinterWrangles II
terion level (often 0.05). The
1.19ttbrUcl'·Z11' .'1· ... rz}'
1
1,
; outcome is classified as show
U .... U1C(.21J,.1.'1· •• Ti}; 11,
1 ing statistical signifi · lCance if the
AlqebCnct, · 2 '1
1
. 11' Co, to, 0, ,}. 11 .
u""0;cl.2I1'.'I'., {o, i, 0, 11, I actual alpha (probability of the
u"',""cl. 2I1'. 'I'., {i, IT' + IT}' 11. I outcome under the null hypoth
.1"""<1>211'. 11' •. {i, Tz ' + T,}' 11} esis) is no greater than this
_ algorithm nominal alpha criterion level.
1. a formal statement, clear This reasoning is applicable for
all types of statistical testing,
complete and unambiguous, of I
how a certain process needs to : including rerandomisation sta
be undertaken. I tistics which are the concern of
2. an algorithm expressed in a I this present glossary.
programming language for a  alphametic
computer . a cryptarithm in which the let
_ allometry I ters, which represent distinct
any change of shape with size. I digits, form related words or
It describes any deviation of the meaningful phrases.
bivariate relation from the  alternate exterior angles
simple functional form y Ix = c, I if two parallel lines are cut by a
where c is a constant x and I transversal, alternate exterior
y are size measures in units of : angles are outside the parallel
the same dimension. I lines and on opposite sides of
 alpha
I the transversal.
also known as size or typel er I  alternate interior angles
ror. This is the probability that,
according to some null hypoth
esis, a statistical test will gener I
ate a falsepositive error : af
firming a nonnull pattern by
chance. Conventional method I
ology for statistical testing is, in
advance of undertaking the
test, to set a nominal alpha cri I
if two parallel lines are cut by a
y
;"';';'z
9
*=================
transversal, alternate interior ~ or the line containing the oppo
angles are inside the parallel ; site side, or the length of the
lines and on opposite sides of : altitude segment.
the transversal. ~ • altitude of a conic solid
• alternative hypothesis ~ the length of a segment whose
in hypothesis testing, a null hy ; endpoints are the vertex and a
pothesis (typically that there is : point on the plane of the base
no effect) is compared with an ~ that is perpendicular to the
alternative hypothesis (typically ~ plane of the base
that there is an effect, or that
there is an effect of a particular
sign). For example, in evaluat
ing whether a new cancer rem
I • altitude of a cylindric solid
~ the distance be,tween the planes
: of the bases
edy works, the null hypothesis ~ • altitude of a trapezoid
typically would be that the rem I the distance between the bases
edy does not work, while the ~ of a trapeziod
alternative hypothesis would be
that the remedy does work.
When the data are sufficiently
improbable under the assump
tion that the null hypothesis is
true, the null hypothesis is re
jected in favour of the alterna
tive hypothesis. (This does not
: • altitude of a triangle
I .
: the perpendicular segment
I from a vertex to the line con
; taining the opposite side of a
: triangle
I
: • ambiguous
~ not stable; changing
imply that the data are probable I • amicable numbers
under the assumption that the
alternative hypothesis is true,
nor that the null hypothesis is
false, nor that the alternative
hypothesis is true.
• altitude
(of a triangle) A line segment
drawn from a vertu that is per
pendicular to the opposite side
; two numbers are said to be
: amicable if each is equal to the
~ sum of the proper divisors of
; the other.
; • analyse
: to break down into parts and
~ explain or demonstrate the
I logic of a situation or a process.
10
angle I anisotropy II
• angle
1. (of a polygon) an angle hav
ing its vertex at one of the
polygon's vertices, and having I
two of the polygon's sides as its
I tex at the viewer's eye, one side
I horizontal, and the viewer's line
of sight to the object as the
other side.
I • angle of rotation
sides.
2. two noncollinear rays (the
sides of the angle) having a com I
mon endpoint (the angle's ver
I the angle between a point and
its image under a rotation, with
its vertex at the center of the
I rotation. and sides that go
tex).
• angle bisector
through the point and its im
age. Also, the measure of the
angle.
I • angle ruler
a ray that has the vertex of the I
angle as its endpoint, and that
divides the angle into two con
gruent angles.
a hinged ruler with a protrac
tor attached for reading the
I measure of an angle in degrees. LI!OH'''' ..\.ooe •• _
I • angle side
one of the two rays forming an
angle
• angular velocity
I (of an object moving around a
circle) the rate of change, with
respect to time, of the measure
• angle of depression I of the central angle that inter
(of an object lower than the I cepts the arc between the ob
viewer) The angle with its ver : ject and a ftxed point.
tex at the viewer's eye, one side I
horizontal, and the viewer's line I • anisotropy
of sight to the object as the anisotropy is a descriptor of one
other side. I aspect of an affine transforma
• angle of elevation
(of an object highet than a
viewer) The angle with its ver
tion. In two dimensions, this is
the ratio of the axes of the el
I lipse into which a circle is trans
formed by an affine transfor
IllInnulus I lire length
11
mation. In general, it is the _ antecedent
maximum ratio of extension of ; the 'if' part of a conditional'
in one to exten ; represented by p; aka hypoth:
m a perpendlCular direc : esis, given, problem
tlon.
 annulus
the region between two concen
tric circles of unequal radius.
I ••
:  antlpnsm
I a polyhedron resulting from
: rotating one base of a prism
and connecting the vertices so
I that the lateral faces are tri
angles.
 apothem
: (of a regular polygon) a line
segment between the center of
; the polygon's circumscribed
: circle to a side of the polygon
that is also perpendicular to
; that side. Also, the length of
: that line segment.
I
:  arc
_ ANSI (of a circle) Two points on the
acronym for the American Na ; circle (the endpoints of the arc)
tiona! Standards Institute. This : and the points of the circle be
body publishes specifications tween them. An angle inter
for a number of standard pro I cepts an arc if the sides of the
gramming languages. The angle the circle at the
specifications are generally ar : of the arc. arc is
ranged to concur with those of mcluded by the chord wlth the
ISO, ; same endpoints.
_ ante  arc length
the upfront cost of a bet: the : the portion of the circumference
money you must pay to play the of the circle described by an arc,
game. From Latin for "before." ; measured in units of length.
MII,.",.riu======= II
12
• arc measure
the measure of the central angle
that intercepts an arc, measured
in degrees.
B
/ ; "
.dAC .Sl..... / "
•. / / '"
",
.ID< I '
;;,dA(  ' .00 ;, "'
I I · ,
\ / _;;0
;/;
'
"' ,
• area
the amount of space taken up
in a plane by a figure
• arithmetic mean
I association, which is a special
I case of association in which
large values of one variable tend
I to occur with large values of the
I other, and small values of one
tend to occur with small values
of the other (positive associa
I tion), or in which large values
of one tend to occur with small
values of the other, and vice
I versa (negative association).
I • associative property
property about grouping of
numbers; of addition, the c for
I mula (a + b) + c = a + (b +
c).
• assume
the arithmetic mean of n num I
bers is the sum of the numbers
divided by n.
to accept as true without facts
I or proof.
• association I • asymptote
two variables are associated if I a straight line always approach
some of the variability of one
can be accounted for by the
other. In a scatterplot of the two
variables, if the scatter in the
values of the variable plotted on
the vertical axis is smaller in
narrow ranges of the variable
plotted on the horizontal axis
(i.e., in vertical "slices") than it
is overall, the two variables are
associated. The correlation co
efficient is a measure of lillear
ing but never intersecting a
curve.
I • automatic drawer
a computer program that lets
you build constructions
• automorphism
an isomorphism from a set
I onto itself.
I • average
a sometimes vague term. It usu
ally denotes the arithmetic
II =======.MJJ,.",.,;u
mean, but it can also denote the
median, the· mode, the geomet
ric mean, and weighted means,
among other things.
13
ther lands heads or lands tails
; is the sum of the chance that the
: coin lands heads and the chance
that the coin lands tails, because
I both cannot occur in the same
: coin toss. All other mathemati
cal facts about probability can
I be derived from these three axi
oms. For example, it is true that
: the chance that an event does
not occur is (100% the chance
; that the event occurs). This is a
: consequence of the second and
• axioms of probability third axioms.
there are three axioms of prob I • axis
ability: Chances are always at . (ofa c lind ) th lin
I Th I Y er, e e segment
east e that : connecting the centers of the
something happen! IS 100%. If I bases
two events cannot both occur at : .
the same time (if they are dis • base
joint or mutually exclusive), the ; I. the side of an isoscoles tri
chance that either one occurs is : angle whose endpoints are the
the sum of the chances that vertices of the base angles
each occurs. For example, con I 2. a side of a polygon or face of
sider an experiment that con : a solid used for reference when
sists of tossing a coin once. The drawing an altitude or other
first axiom says that the chance I feature.
the coin lands heads, for 3. a face of a solid used for ref
mstance, must be at least zero. : erence when drawing an alti
The second axiom says that the I tude or other feature.
chance that the coin either lands ; 4. the congruent parallel poly
heads or lands tails or lands on : gons of a prism. If the faces are
its edge or doesn't land at all is all rectangular, any parallel pair
100%. The third axiom says; can be considered the bases.
that the chance that the coin ei
14
Otherwise, the nonrectangular P(AIB) = P(B IA) xP(A)/(
pair forms the bases. i P(B I A) xP(A) + P(B lAc)
5. in the expression xY, x is called : xP(Ac».
the base and y is the exponent. • bending energy
• base angles bending energy is a metaphor
1. (of an isosceles triangle) The i borrowed for use in
two angles opposite the two : morphometrics from the me
congruent sides. chanics of thin metal plates.
2. (of an isosceles triangle) The I Imagine a configuration of
two angles opposite the two : landmarks that has been printed
congruent sides. on an infinite, infinitely thin,
3. (of a trapezoid) A pair of I flat metal plate, and suppose
angles with a base of the trap that the differences in coordi
ezoid as a common side. : nates these same ¥dmarks
• b Ii I in ano er picture are taken as
ase ne : . A: I f tho
for a system of twopoint shape I vernc o. IS
din t
r. I dm ks· : plate perpendicular to Itself,
coor a es lor an ar m a Ie· dina
I th b lin
. th lin one artesIan coor te at a
pane, ease elS e e:. Th be din f
connecting the pair of land I nme. eng energy 0
th
. d 1::._ d : one of these outofplane
madcs at are asslgne to llAe I" h h )). th (·deal
1
'. (0 0) d (1 0)· th . s ape c anges IS e); 
ocanons , an , me· ized) th uld be
. I I I energy at wo re
constructIon. n genera,. . ed be d th tal I
baselines work better if they are i qwr to n e me pate
closely aligned with the long : s.o that the landmarks .were
. f I lifted or lowered appropnately.
aXIS 0 the mean landmark . Whil· h· bending
shape and pass near the centroid i. eeal
m
p
f th t h . IS a r quannty, measur m
o a mean s ape ; appropriate units (g cm2 sec
2
),
• Bayes' rule there an alternate formula
Bayes' rule expresses the con : that remains meaningful in
ditional probability of the event morphometries: bending en
A given the event B in terms of i ergy is proportional to the in
the conditional probability of : tegral of the summed squared
the event B given the event A: second derivatives of the "ver
II hentlinamewmlltriJ&lhettl
tical" displacement the extent to energy of a general transforma
which it varies from a uniform; tion is the sum xtLklX +y'Lk:ly
tilt. The bending energy of a of bending energy of Its
shape change is the sum of the : honzontal mod
bending energies that apply to I eled as vemcal
any two perpendicular coordi the bending energy of Its
nates in which the metaphor is : cal ycomponent, modeled Slffil
evaluated. The bending energy I !arly as a "vertical" plate.
of an affme transformation is • Bernouilli process
zero since it corresponds to a : this is the simplest probability
tilting of the plate with?ut any model a single trial between
bending. The value obtamed for I two possible outcomes such as
the bending energy correspond : a coin toss. The distribution
. ing to a given is depends upon a . single
inversely pr?pomonal to scale. parameter,'p', representmg the
quantities should not ?e ; probability attributed to one
interpreted as measures dis : defmed outcome out of the two
similarity (e.g:, taxonOIll1C or possible outcomes.
evolutionary distance) between .
• beta
two forms. .' 2 be
I also known as type error, ta
• bending energy : is the complement to power :
the formula for bending beta = (Ipower). This is the
the formula whose IS pro ; probability that a statistical
portional to that mtegral of : will generate a falsenegative
those summed squared. second error: failing to assert a defmed
derivatives is. a quadratic form ; pattern of deviation from a null
(usually wntten Lk·l) deter : pattern in circumstances where
mined by the coordinates of the the defmed pattern exists. Con
landmarks of the reference I ventional methodology for sta
form. That is, if h is a vector : tistical testing is to set in ad
describing the heights of a plate vance a nominal alpha criterion
above a set of landmarks, then I level the corresponding level for
bending energy is htLklh. In : BETA will depend upon the
morphometrics, the bending nominal alpha criterion level
I
Msth_tics====== II
16
bias I biMr,ynumber II
and upon further considerations J swer that differs from the truth.
including the strength of the J The bias is the average (ex
pattern in the data and the pected) difference between the
sample size. Interest is gener J measurement and the truth. For
ally in the relative power of dif J example, if you get on the scale
ferent tests rather than in an with clothes on, that biases the
absolute value. It is question : measurement to be larger than
able whether the concept of J your true weight (this would be
BETA error is properly appli : a positive bias). The design of
cable without considering the an experiment or of a survey can
concept of sampling from a J also lead to bias. Bias can be
population, which is separate deliberate, but it is not neces
from the concerns of this Glos sarily so.
sary. Applicability of this rea I • biconditional
soning is also closely bound up I Q conditional and its converse
with the choice of test statistic. r where the converse is also true;
: (lao::: :=:=:: .. , uses the words if and only if.
..,... ,_Wool ,
t:'416:1 rt., 0..__ I
• _&1Jtt
. .
, •• m I
IAJ pI"D ","" , I
toueo. ....,,'.. I
rd'fr.l lAl .p} ... ,l p .... ) l '.' •• I
,.",... .... ""''''O''.I0(h3 t14tiIJ , ::=::.:r ......... :
• bijection
I a onetoone onto function.
I • bilateral symmetry
..... a,t"";""IJ.Q:d .. ,.th.
i=:=::';:;;; .. ••• ......... __
::::1Ia.J WltS)
I reflectional symmetry with only
one line of symmetry.
I
....... ,,.. ...... )., ih:::::" 1
.... uacD
::::= It<W1 I
= ,\I)
ttel!'" a:u..
• bimodal
I having two modes.
I • binary number
r ......  =.1
:=
I a number written to base 2.
...  5..,=r.a
1loCt... IiaJtJalnn r.a.,os.'
• bias
a measurement procedure or J
estimator is said to be biased if, I
on the average, it gives an an
_ ..
...... "'.add
..."""
I;' :.
32
.. 
_ ... 2
:.
.... • tIII OIl
..
•
2
•
..
•
.. ..
.
II hiM", operation I biseet
17
• binary operation probability histogram of the
a binary operation is an opera I binomial distribution.
tion that involves two operands.
F
; • binomial test
or example, addition and sub. .. . . .
traction are binary operations. ; this IS a test refemng
•. . : to a repeated bmary process
I such as would be expected to
an expression that IS the sum of : generate outcomes with a bino
two terms. mial distribution. A value for
• binomial coefficient the parameter 'p' is
the coefficients of x in the ex ; hypothesised (null hypothesis)
pansion of (x+ 1 )n. : and the difference of the actual
I value from this is assessed as a
• binomial distribution
; value of alpha.
a random variable has a bino
mial distrib,:!:tion (with param • biplot
eters n and p) if it is the num : a single diagram that represents
ber of "successes" in a fixed two separate scatterplots on the
number n of independent ran ; same pair of axes. One scatter
dom trials, all of which have the : is of some pair of columns of
same probability p of resulting the matrix U of the singular
in "success." Under these as I value decomposition of a ma
sumptions, the probability of k : trix S, and the other scatter is
successes (and n _ k failures) is of the matching pair of columns
nC
k
pk(lp)nk, where nC
k
is the I of V. When S is a centered data
number of combinations of n the effect is to plot prin
objects taken k at a time: nC
k
= : Clpal component loadings and
n!j(k!(nk)!). The expected scores on the same diagram.
value a • biquadratic equation
the Bmomial dIstnbutIon IS ; a polynomial equation of the
nxp, and the.stanruu:d error of : 4th degree.
a random vanable With the Bi I •
nomial distribution is (nxpx(l .
p» V2. This page shows the ; to diVide mto two congruent
I parts.
18
I bootstmp II
• bisector pf a segment human subjects, it is usually
any plane, point or twodimen ; necessary to administer a pla
sional figure containing the the cebo to the control group.
midpoint of the segment and no I
• bootstrap
other points on that segment I this is a form of randomisation
• bit I test which is one of the alterna
a binary digit. tives to exhaustive re
• bivariate randomisation. bootstrap
having or having to do with two I scheme involves generating
subsets of the data on the basis
variables. For example, bivari I
ate data are data where we have of random sampling with re
two measurements of each "in I placements as the data are
dividual." These measurements I sampled. Such resampling pro
might be the heights and vides that each datum is equally
weights of a group of people I represented in the
(
". di ·dual"· ) th I randomisation scheme; how
VI
f
IS a person, e: ever, the bootstrap procedure
hel ts 0 fathers and sons (an I h Co hi h di· . h
". d··d 1"· f h . as leatures w c songws
III IVI ua IS a at erson .. f h d f
.) th d I It rom t e proce ure 0 a
pair, e pressure an tem. Mel Th di .
. onte ar 0 test. e stm
perature of a fIXed volume of I ..
(
". di . dual"· th 1· gwshing features of the boot
gas an
f
ID VId IS vo  ; strap procedure are concerned
ume 0 gas un er a certam set. . th lin th .
of experimental conditions), ; WI th
g
erebels nO
f
S 1 th 1
·· constramt upon e num r 0
etc. catterp ots, e corre anon·. tha d be
coefficient, and regression nmes d a atum inglre
p

make sense for bivariate data ; resente lingID genberaon
th
g
a e
f
b
.. d . resamp su set; e SIZe 0
ut not uruvariate ata. ; th lin b be
e resamp g su sets may
• blind experiment fixed arbitrarily independently
in a blind experiment, the sub : of the parameter values of the
jects do not know whether they experimental design and may
are in the treatment group or ; even exceed the total number of
the control group. In order to : data. The positive motive for
have a blind experiment with bootstrap resampling is the gen
1I=======MAthemsties
II bootstrapestimateofstandarderror I : model 19
eral relative ease of devising an the sample to estimate the SE
appropriate resampling algo ; of sampling from the popula
rithm when the experimental : tion. For sampling from a box
design is novel or complex. A of numbers, the SD of the
negative aspect of the bootstrap ; sample is the bootstrap estimate
is that the form of the : of the SD of the box from
resampling distribution with which the sample is drawn. For
prolonged resampling con I sample percentages, this takes
verges to a form which depends a particularly simple form: the
not only upon the data and the : SE of the sample percentage of
test statistic, but also upon the I n draws from a box, with re
bootstrap resampling subset placement, is SD(box)/nY2,
size thus the resampling distri : where for a box that contains
bution should not be expected only zeros and ones, SD(box)
to converge to the gold standard ;  ( (fraction of ones in
form of the exact test as is the : box) x (fraction of zeros in box)
case for MonteCarlo )Y2. The bootstrap estimate of
resampling. An effective neces ; the SE of the sample percent
sity for the bootstrap procedure : age consists of estimating
is a source of random codes or SD(box) by ((fraction of ones
an effective pseudorandom I in sample) x (fraction of zeros
generator. : in sample»'h. When the
sample size is large, this ap
 bootstrap estimate of
standard error I proximation is likely to be
the name for this idea comes good.
from the idiom "to pull oneself :  box
up by one's bootstraps," which a surface made up of rect
connotes getting out of a hole angles; a rectangular parallel
without anything to stand on. ; epiped
The idea of the bootstrap is to
assume, for the purposes of es
timating uncenainties, that the
sample is the population, then
use the SE for sampling from
• box model
: an analogy between an experi
ment and drawing numbered
; tickets "at random" from a box
: with replacement. For example,
I
II
20
, al I
suppose we are trymg to ev u ,
ate a cold remedy by giving it ;
or a placebo to a group of n in :
dividuals, randomly choosing
half the individuals to receive I
the remedy and half to receive :
the placebo. Consider the me
dian time to recovery for all the I
individuals (we assume every
one recovers from the cold :
M
R
Q
p
N
eventually; to simplify things, • box plot
we also that no one ,re ; a representation of data above
covered m exactly the median: a numbered scale where the
time, and that n is' even). By "box" encloses all data between
defmition, half the individuals the median of the lower half
got better in less than, the me : (quartile 1) and the median of
dian time, and half 10 more the upper half (quartile 3), with
than the median time. The in I a vertical line inside the box to
dividuals who received the: indicate the median of the data;
treatment are a random sample a dot represents each of the high
of size n/2 from the set of n I and low values of the data, and
subjects, half of whom got bet a horizontal line called a whis
ter in less than median time, and : ker connects each dot to the
half in longer than median box.
time, If the remedy is ineffec; b h d b d
. f b' • ranc an  oun
tIve, the number 0 su Jects :, . ,
h
'd th d d I exploratIon of a randOlrusanon
w 0 recelve e reme y an ",'
ho d
' 1 th : distrlbunon m such a way as to
w recovere mess an me I ' ,
di
' 'lik th f nl2 . anOclpate the effect of the next
an orne lS e e sum 0, "I' th
' fi I rannOffilSatIOn re atIve to e
draws Wlth replacement rom a , d ' . This al
box with two tickets in it: one ; present OffilSanon. ,
with a "1" on it and one with a : lows selecove search of
"0"" I lar zones of a randoffilsatIon
on It. . dis 'b' . th f
. trl uoon; m e context 0 a
randomisation test such selec
II =======MsthtmUJms
II hreaItdownpoint I canonicalcorreUJ";"nalysis 21
tive search may be concerned ~ relations. Each score (linear
with the tail of the i combination) from either list is
randomisation distribution : correlated with no other com
~ bination from its list and with
• breakdown point .
the breakdown point of an esti ~ ~ n l y one score from the other
mator is the smallest fraction of i st.
observations one must corrupt
to make the estimator take any
value one wants.
• byte
the amount of memory needed
to represent one character on a
computer, typically 8 bits.
• calculator notation
the symbols used by a calcula
tor for scientific notation.
• caliban puzzle
a logic puzzle in which one is
asked to infer one or more facts
from a set of given facts.
• canonical
'
~ • canonical correlation
i analysis
: a multivariate method for as
; sessing the associations be
I tween two sets of variables
: within a data set. The analysis
~ focuses on pairs of linear com
I binations of variables (one for
~ each set) ordered by the mag
: nitude of their correlations with
~ each other. The fIrst such pair
; is determined so as to have the
: maximal correlation of any
~ such linear combinations. Sub
; sequent pairs have maximal cor
: relation subject to the constraint
~ of being orthogonal to those
I previously determined.
a canonical description of any
statistical situation is a descrip
tion in terms of extracted vec
tors that have especially simple
ordered relationships. For in
stance, a canonical correlations
analysis describes the relation
between two lists of variables
in terms of two lists of linear
combinations that show a re
markable pattern of zero cor I
Msthmulties======= II
22 canonical 11am; analysis I causation, causal relation II
• canonical variates analysis capacity
a method of multivariate analy ; the amount of liquid that can fill
sis in which the variation among : an object.
groups is expressed relative to I • I
the pooled withingroup cova I • cartesian p ane .
. . C . I . a rectangular coordinate svstem
nance matrIX. anoruca varI . . . J'
al . finds lin I COnSIStIng of a honzontal num
fi
ates an. YSIS ear ber line (xaxis) and a vertical
ormatlons of the data which I b lin ( .).
maximise the among group ethr e . , mtersect
h

. ., I mg at e ongm (zero on eac
varIatIOn relatIve to the pooled . be lin)
withingroup variation. The ca ; num r e.
nonica! variates then may be I • categorical variable
displayed as an ordination to : a variable whose value ranges
show the group centroids and over categories, such as {red,
scatter within groups. This may ; green, blue}, {male, female},
be thought of as a "data reduc : {Delhi, Calcutta, Mumabai},
tion" method in the sense that {short, tall}, {Asian, Mrican
one wants to describe among ; American, Caucasian, Hispanic,
group differences in few dimen : Native American, Polynesian},
sions. The canonical variates are {straight, curly}, etc. Some cat'
uncorrelated, however the vec ; egorical variables are ordinal.
tors of coefficients are not or : The distinction between cat
thogonal as in Principal Com egorical variables and qualita
ponent Analysis. The method is I tive variables is a bit blurry.
closely related to multivariate .
I • catenary
analysis of variance : a curve whose equation is y =
(MANOVA), multiple discrimi (a/2)(e
X
/
3
+e
X
/
3
). A chain sus
nant analysis, and canonical cor I pended from two points forms
relation analysis. A critical as : this curve
sumption is that the within •.
group variancecovariance;. causation, causal relation
structure is similar otherwise : two variables are causally re
the pooling of data over lated if changes in the value of
groups is not very sensible. i one cause the other to change.
For example, if one heats a rigid
II =======.M.thelulies
II ceilingfunction I centroid
container filled with a gas, that
causes the pressure of the gas
in the container to increase.
Two variables can be associated
without having any causal rela
tion, and even if two variables
have a causal relation, their cor
relation can be small or zero.
• ceiling function
the ceiling function of x is the
smallest integer greater than or
equal to x.
• center
(of a circle or sphere) The point
from which all points on the fig
ure are the same distance.
• center of a circle
the point that all points in the
circle are equidistant from.
.. \
'''' ... \
• center of a rotation
the point where the two inter
secting lines of a rotation meet
23
*=================
I • center of gravity
I the mean of the coordinates of
points in a figure, whether one,
I two, or threedimensional
I • central angle
I (of a chord or arc) An angle
: whose vertex is the center of a
I circle and whose sides pass
I through the endpoints of a
chord or arc.
• central angle of a circle
an angle whose vertex is the
I center of the circle
I • central limit theorem
the central limit theorem states
that the probability histograms
I of the sample mean and sample
sum of n draws with replace
ment from a box of labeled tick
I ets converge to a normal curve
I as the sample size n grows, in
the following sense: As n grows,
I the area of the probability his
I togram for any range of values
approaches the area tmder the
normal curve for the same
range of values, converted to
standard units.
• centroid
the point of concurrency of a
I triangle's three medians.
MlJthemR.tiu======= II
",,24=========== .. centroid size I Chebychev'si1UlJUR1ity II
_ centroid size
 cevian
centroid size is the square root I
of the sum of squared distances
of a set of landmarks from their
a line segment extending from
a vertex of a triangle to the op
I posite side.
centroid, or, equivalently, the I
square r'?Ot of the sum of the ~
variances of the landmarks
about that centroid in xand y I
directions. Centroid Size is used
in geometric morphometries _ chance variation, chance
because it is approximately I error
uncorrelated with every shape a random variable can be de
variable when landmarks are composed into a sum of its ex
distributed around lTiean posi ~ pected value and chance varia
tions by independent noise of ; tion around its expected value.
the same small variance at ev The expected value of the
ery landmark and in every di I chance variation is zero; the
rection. Centroid Size is the size I standard error of the chance
variation is the same as the stan
dard error of the random vari
ablethe size of a "typical" dif
ference between the random
variable and its expected value
 Chebychev's inequality
measure used to scale a con
figuration of landmarks so they
can be plotted as a point in I
Kendall"s shape space. The de
nominator of the formula for
the Procrustes distance between I
two sets of landmark configu
rations is the product of their
Centroid Sizes.
I for lists: For every number
k>O, the fraction of elements
in a list that are k SD's or fur
I ther from the arithmetic mean
 certain event
an event is certain if ItS prob
ability is 100%. E v e ~ if an event
is certain, it might not occur.
However, by the complement
rule, the chance that it does not
occur is 0%.
of the list is at most Ijk2. For
random variables: For every
I number k>O, the probability
I that a random variable X is k
SEs or further from its expected
I value is at most l/k2.
II chisquare curve I chisquared 25
• chisquare curve dependent trials, each of
the chisquare curve is a fam ; which can result in one of k
ily of curves that depend on a : possible outcomes. Suppose
parameter called degrees of that in each trial, the probabil
freedom (d.f.). The chi ; itythatoutcome i occurs is pi,
square curve is an approxima : for i = 1, 2, ... , k, and that
tion to the probability histo these probabilities are the
gram of the chisquare statis I same in every trial. The ex
tic for multinomial model if : pected number of times out
the expected number of out come 1 occurs in the n trials
comes in each category is I is n.xp1; more generally, the
large. The chisquare curve is expected number of times out
positive, and its total area is : come i occurs is
100%, so we can think of it as • chisquared distribution
the probability histogram of a where expected frequencies are
random variable. The balance
point of the curve is d.f., so
the expected value of the cor
responding random variable
would equal d.f .. The stan
dard error of the correspond
ing random variable would be
(2xd.f.)V2. As d.f. grows, the
shape of the chisquare curve
approaches the shape of the
normal curve.
; sufficiently high, hypothesised
: distributions of counts may be
approximated by a normal dis
; tribution rather than an exact
: binomial distribution. The cor
responding distribution of the
; chisquared statistic can be de
: rived algebraically this is the
chisquared distribution which
I has been computed and pub
historically as extensive
• chisquare statistic : prmted tables. Use of the tables
the chisquare statistic is used is notably simple, as the chi
to measure the agreement be ; squared distribution depends
tween categorical data and a : upon only one parameter, the
multinomial model that pre degrees of freedom, defined as
diets the relative frequency of ; one less than the number of cat
outcomes in each possible cat : egories.
egory. Suppose there are n in I
Mslthema.ties======= II
26 chiSlJ.tl4red statistic I circumfoYence II
• chisquared statistic I Chord
this is a longestablished test c
statistic for measuring the ex Q
tent to which a set of categori :
D "A
cal outcomes depart from a. I
hypothesised set of probabili E F
ties. It is calculated as a sum of : A line segment that connects 2
terms over the available catego points on a circle.
ries, where each term is of the i CD and EF are chords of circle A.
form: ((0E)2)fE; '0' repre • circle
sents the observed frequency for I the set of points on a plane at a
the category and 'E' represents certain distance (radius) from
the corresponding a certain point (ceI1ter); a poly
quency based upon multIplymg : gon with infinite sides
the sample size by the I •
h thesised probability for the : • cl1'cular cone . .
being considered a cone whose base IS a crrcle.
(therefore 'E' will generally not I • circularity
be an integer value). In situa : when on a search, circling back
tions where the of cat to a previous place visited (defi
egories is 2 an alternatIve pro nition, web site, etc.), ustially
cedure is to use an exact i unhelpful or redundant
biniomial test.
i • circumcenter
• chord : the circumcenter of a triangle
1. a line segment whose end is the center of the circum
points lie on a circle.. ; scribed circle.
2. the line joining two pomts on
a curve is called a chord.
• chord of a circle
a segment whose
are on a circle
i • circumcircle
: "the circle circumscribed about
I
: a figure.
• circumference
i the distance around a circle,
given by the formula C = 2m;
II circumscribed I cluster analysis
27
*===============
where r is the radius of the • class interval
circle. ; in plotting a histogram, one
• circumscribed : starts by dividing the range of
passing through each vertex of into a set of
a figure, usually referring to pmg called class mter
circles circumscri bed around . vals, m such a way that every
polygons or spheres circum _ is contained in some class
scribed around polyhedrons. I mterval.
The figure inside is inscribed in I • class of functions
the circumscribed figure. family of functions such as lin
• cissoid : ear, quadratic, power (polyno
a curve with equation y2(a_ mial), exponential, or logarith
X)=X3. ; mic.
P2
• class boundary
a point that is the left endpoint
of one class interval, and the
right endpoint of another class
interval.
• classes of numbers
: family of numbers or number
systems such as natural, inte
; ger, rational, irrational, real, or
: complex.
I
: • classify
to categorise something accord
; ing to some chosen character
: istics.
I
: • clockwise
in orientation, the direction in
; which the points are named
: when, if traveling along the line,
the interior of the polygon is on
the right.
I • cluster analysis
a method of analysis that rep
: resents multivariate variation in
I data as a series of sets. In biol
ogy, the sets are often con
II
28
corJJicient I combinatitms II
structed in a hierarchical man ~ • coincidental lines
ner and shown in the form of a ; lines that are identical (one and
treelike diagram called a den : the same)
drogram. ~ • collinear
• coefficient ~ lying on the same line.
a coefficient, in general, is a ~ • combinations
number multiplying a function.
In multivariate data analysis,
usually the "function" is a vari I
able measured over the cases of :
the analysis, and the coefficients I
multiply these variable values I
before we add them up to form
the number of combinations of
n things taken k at a time is the
number of ways of picking a
subset ofk of the n things, with
out replacement, and without
regard to the order in which the
elements of the subset are
pickc:d. The number of such
combinations is nC
k
= n!j(k!(n
a score. A coefficient is not the
same as a loading.
• coincide k)!), where k! (pronounced "k
lying exactly on top of each factorial") is kx(kl)x(k2)x
other. Line segments that coin ... x 1. The numbers nC
k
are
cide are identical; they have all I also called the Binomial coeffi
the same points. I cients. From a set that has n el
ements one can form a total of
I 2n subsets of all sizes. For ex
I ample, from the set {a, b, c},
: which has 3 elements, one can
I
: form the 2
3
= 8 subsets U, {a},
I {b}, {c}, {a,b}, {a,c}, {b,c},
{a,b,c}. Because the number of
subsets with k elements one can
I form from a set with n elements
is nc
k
, and the total number of
subsets of a set is the sum of
~ the numbers of possible subsets
; of each size, it follows that
nco +nC
1
+nC
2
+ ... +nC
n
= 2n.
II
The calculator has a button  complementary angles
(nc
m
) that lets you compute the ; two angles whose measures
number of combinations of m : have the sum 90°.
things chosen from a set of n I
things. To use the button, first I
type the value of n, then push
the nC
m
button, then type the
value of m, then press the "=" I
button.
ArSfc.
8 ()
mLABC + mLCBD = 90"
 commutative properties
properties about order of addi
tion' a + b = b + a; of multi
plication, a x b = b x a"
 compass
a drawing tool used to draw
circles at different radii
 compatible numbers
numbers that can be easily ma
nipulated and operated on men
tally.
LABC and LCBD are complementary angles.
I
:  complex numbers
complex numbers are an alge
; braic way of coding points in
: the ordinary Euclidean plane
so that translation (shift of
; position) corresponds to the
: addition of complex numbers
I .
: and both rescaling (enlarge
I ment or shrinking) and rota
tion correspond to multiplica
: tion of complex numbers. In
_ complement rule this system of notation, in
the probability of the comple ; vented by Gauss, the xaxis is
ment of an event is 100% mi : identified with the "real num
nus the probability of the event: bers" (ordinary decimals
P(Ac) = 100% peA). ; numbers) and the yaxis is
I
: identified with "imaginary
 comp ement I b" ( h f
th I f b f . num ers t e square roots 0
e.comp 0 a su .set 0 ; negative numbers). When you
a gIven set IS the collecnon of" ul" I" thi" b
all elements of the set that are ; tip r pomts on
d
" s axiS h
Y
not elements of the subset. : t emse ves accor mg to t e
I rules, you get negative points
: on the "real" axis just defined.
Many operations on data in
MJJ.th_tics========== II
/I
two dimensions can be proved I • concentric circles
valid more directly if they are I circles that share the same
written out as operations on center, but have different ra
complex numbers. I dii
• compose numbers
put a set of numbers together I
to form a new number using
addition or multiplication.
• composite transforma
tion
the composite of a first trans
formation S and a second
transformation T is the trans I
formation mapping a point P
onto T(S(P».
• concrete materials
• composition
I objects to be manipulated (e.g.,
pattern blocks, snap cubes,
geoboards, tangrams, color
tiles, base ten blocks).
(of transformations) The trans I
formation that results when
one transformation is applied
after another transformation.
• concurrent
• compound eventS
two or more events in a prob
ability situation such as flipping
a coin and spinning a spinner.
• concave
curved from the inside.
• concave polygon
a polygon having at least one
diagonal lying outside the poly
gon; not convex.
I intersecting at a single point
( called the point of
I concurrency).
I • conditional
I a statement that tells if one
thing happens, another will fol
low.
• conditional probability
I suppose we are interested in the
probability that some event A
occurs, and we learn that the
I event B occurred. How should
II =======MRthem4ries
we update the probability of A
to reflect this new knowledge?
This is what the conditional
probability does: it says how the
additional knowledge that B
occurred should affect the prob
ability that A occurred quanti
tatively. For example, suppose
that A and B are mutually ex
clusive. Then if B occurred, A
did not, so the conditional prob
ability that A occurred given
that B occurred is zero. At the
other extreme, suppose that B
is a subset of A, so that A must
occur whenever B does. Then
if we learn that B occurred, A
must have occurred too, so the
conditional probability that A
occurred given that B occurred
is 100%. For inbetween cases,
where A and B intersect, but B
is not a subset of A, the condi
tional probability of A given B
is a number between zero and
100%. Basically, one "restricts"
the outcome space S to consider
only the part of S that is in B,
because we know that B oc
curred. For A to have happened
given that B happened requires
that AB happened, so we are
interested in the event AB. To
have a legitimate probability
requires that P(S) = 100%, so
31
*=================
~ if we are restricting the out
; come space to B, we need to
: divide by the probability of B
~ to make the probability of this
; new S be 100%. On this scale,
: the probability that AB hap
~ pened is P(AB)fP(B). This is
, the deftnition of the conditional
: probability of A given B, pro
~ vided P(B) is not zero (division
~ by zero is undefined). Note
; that the special cases AB = {}
: (A and B are mutually exclu
~ sive) and AB = B (B is a subset
; of A) agree with our intuition
: as described at the top of this
~ paragraph. Conditional prob
; abilities satisfy the axioms of
: probability, just as ordinary
~ probabilities do.
~  conditional proof
; a proof of a conditional state
: ment.
,
:  cone
~ a solid whose surface consists
, of a circle and its interior, and
: all points on line segments that
~ connect points on the circle to a
~ single point (the cone's vertex)
; that is not coplanar with the
: circle. The circle and its interior
~ form the base of the cone. The
; radius of a cone is the radius of
: the base. The altitude of a cone
Msthemtllti&s========== II
II
is the line segment from the the family, and these likelihoods
vertex to the plane of the base ; may be then used to defme a
to it. The contiguous set of values which
of a cone 18 t.he length of : occupy a certain proportion of
Its If the line segment I the total unit weight of the like
the vertex of a cone : lihoods integrated over all val
with center of its base is ues of the test statistic, the con
the base, then fidence interval is defmed by
the cone 18 a nght cone; other ' the minimum and maximum
wise it is oblique. values of the range of values so
• confidence interval defined. The proportion of the
for a given rerandomisation ; total weight within the range of
distribution, a family of related : values is regarded as an alpha
distributions may be defined the of the
according to a range of hypo I test stanstic hes WIthin this
values of the pattern range, the defmition
whIch the test statistic mea : of a confidence mterval cannot
sures. For instance, for the pit be unique imposing
man permutation test to test for ' further constramts. Approaches
a scale shift between .two to providing suitable con
groups, a related distribution that a confidence
may be formed by shifting all ; mtet;al will be unique, include
the observations in one group : the confidence interval
by a common amount where : to mclude the whole of one tail
this common shift is ; of the or to be
as a continuous variable, With : centred m some sense upon the
fmite numbers of data the num outcome value; or to be centred
ber of related distributions will ; bet,ween TAILS of equal
be fmite, and typically consid weIght" In, the ,cas,e of re
erably smaller than the number : randomIsanon distributions,
of points of the randomisation I these are discrete distributions
distribution. The likelihood of : so there will generally be no
the outcome value may be cal range of,values with weight cor
culated for each distribution in I responding exactly to an arbi
: trary nominal alpha criterion
II ===:========Mstht:'l'/Ul.tU:s
II con.fideme level I congruentfiDU1'eS. 33
level, and the problem of non the treatment (if any). For ex
uniqueness is therefore not gen ; ample, prominent statisticians
eratly solvable. : questioned whether differences
between individuals that led
; some to smoke and others not
: to (rather than the act of smok
ing itself) were responsible for
I the observed difference in the
frequencies with which smok
: ers and nonsmokers contract
I various illnesses. If that were
the case, those factors would be
 confidence level : confounded with the effect of
the confidence level of a confi smoking. Confounding is quite
dence interval is the chance that ; likely to affect observational
the interval that will : studies and experiments that
data are collected wIll contam I are not randomised Confound
the corresponding If ing tends to be by
one computes confidence mter : randomisation.
vals again and again from inde
pendent tiata the longterm .  congruent
limit of the fr;ction of intervals equilateral, equal, exactly the
that contain the parameter is I same (size, shape, etc.)
the confidence level.
• confounding
when the differences between
the treatment and control
groups other than the treat
ment produce differences in re
sponse that are not distinguish
able from the effect of the treat
ment, those differences be
tween the groups are said to be
confounded with the effect of
;  congruent angles
: two or more angles that have
the same measure.
I • I
:  congruent clrc es
I two or more circles that with the
: same radius.
I
:  congruent figures
two figures where one is the
I image of the other under a re
flection or composite of reflec
: tions.
MIIthemtlnes======= \I
tpolygOllS I constantofRnetpmtUm II
• congruent polygons
• consecutive vertices
(of a polygon or polyhedron),
two vertices that are connected
two or more polygons with the I
exact same size and shape.
I by a side or edge.
• congruent segments
two or more segments that have
the same measure or length.
• conic section
I • consensus configuration
the cross section of a right cir
cular cone cut by a plane. An el
lipse, parabola, and hyperbola I
I a single set of landmarks in
tended to represent the central
tendency of an observed sample
I for the production of superimpo
sitions, of a weight matrix, or
some other morphometric pur
I pose. Often a consensus configu
are conic sections.
1\
/ \!
•...... : ....... .... ..
; ............ , ..
 ... , .....
I =: .. :\ I .. 
='= I \  ) '/
:.,l_tt.. i 1\
I \ l ' \
I \ \
f, .\ / \
( " \
.'/  '/
• conic solid
ration is · computed to optimize
some measure of fit to the full
I sample: in particular, the
Procrustes mean shape is com
puted to minimise the sum of
I squared Procrustes distances
I from the the consensus land
marks to those of the sample.
the set of points between a point I
(the venex) and a noncoplan;:lI
region (the base), including the
point and the region.
• consequent
I the second or "then" part of a
I conditional statement.
• conjecture
a guess, usually made as a re
sult of inductive reasoning.
I • constant of an equation
the term that has no variable in
an equation; example: "0
• consecutive angles
( of a polygon), two angles that I
have a side of the polygon as a I
common side.
• consecutive sides
(of a polygon), two sides that I
have a common vertex.
1 (v  va) 2
x  Xo = vot +  t
2 t
1 1
x  Xo = Vat + tit  vat
2 2
1 1
x  Xa = .,..vat + vt
2 2
1
x  Xo = (tJo + tI)t
2
II =====.===MsJthmuJnes
II 35
• constant rate of change The expected number of suc
set .of data or table of values in ; cesses is np = 10, and the stan
which amount of the depen dard error is (np( Ip» lh = 6lh
dent vanable changes by a con . = 2.45. If we consider the area
stant (fixed) value as the value ; under the normal curve at the
of the independent variable : point 10 successes transformed
changes by a constant value. to standard units,' we get zero:
• construct the area under a point is always
create a figure using only a ; zero. We get a better approxi
straight edge and compass. : mation by considering 10 suc
cesses to be the range from 9
• construction ; 1/2 to 101/2 successes. The only
a precise way of drawing which : possible number of successes
allows only 2 tools: the straight between 91/2 and 10 1/2 is 10,
edge and the compass ; so this is exactly right for the
• continuity correction : binomial distribution. Because
in using the normal approxima the normal curve is continuous
tion to the binomial probabil ; and a binomial random variable
ity histogram, one can get more : is discrete, we need to "smear
accurate answers by fmding the out" the binomial probability
area under the normal curve ' over an appropriate range. The
corresponding to halfintegers, ; lower endpoint of the range, 9
transformed to standard units. : 1/2 successes, is (9.5 10)/2.45
This is clearest if we are seek '. = 0.20 standard units. The
ing the chance of a particular upper endpoint of the range, 10
number of successes. For ex : 1/2 successes, is (10.5 10)/2.45
ample, suppose we seek to ap = +0.20 standard units. The
proximate the chance of 10 suc ; area under the normal curve
cesses in 25 independent trials, : between 0.20 and +0.20 is
each with probability p = 40% about 15.8%. The true bino
of success. The number of suc ; mial probability IS
cesses in this scenario has a bi : 25CI0x (0.4)10x (0.6)15 =
nomial distribution with pa 16%. In a similar way, if we
rameters n = 25 and p = 40%. ' seek the normal approximation
: to the probability that a bino
,
cuntinUIIIIS tlistriburion I cuntmaitm II
I ..
36
mial random variable is • the for any possible of
range from i successes to k suc ; cumulative there IS
cesses, inclusive, we should find an exact; value of
the area under the normal curve : the statIStiC m question.
from iI/2 to k+ 1/2 successes, I • continuous variable
transformed to .units. a quantitative variable. is con
If we seek the probability of : tinuous if its set of pOSSible val
more than i successes and fewer ues is uncountable. Examples
than k successes, we should find ; inclUde temperature, exact
the area under ¢e normal curve : height exact age (including
corresponding to the range i + 1/ parts of a second). In practice,
2 to kI/2 successes, trans lone can never measure a con
formed to standard units. If we : tinuous variable to infinite pre
seek the probability of more cision so continuous variables
than i but no more than k suc ; are approximated by
cesses, we should find the area: discrete variables. A random
, I .
unde.r the normal curve corre : variable X is also called conttnu
sponding to the range i + 1/2 to I ous if its set of possible values
k+ 1/2 successes, transformed is uncountable, and the chance
to standard units. If we the : that it takes any particular value
probability of at least I but is zero (in symbols, ifP(X = x)
fewer than k successes, we ; = 0 for every real number x).
should fmd the area under the : A random variable is continu
normal curve corresponding to ous if and only if its cumulative
the range iI/2 to kI/2 suc ; probability distributi?n function
cesses, transformed to standard : is a continuous funCtIon (a func
units. Including or excluding the tion ,with no jumps).
halfinteger ranges at the ends I •
of the interval in this manner is : • . .
called the continuity correction. the of a
. ure resulttng from multlplymg
• continuous distribution all dimensions of a given figure
a probability distribution of a ; by a number betwec:n zero and
continuous statistic, based upon one.
an algebraic formula, such that :
I
37
II contm,positi1le I cunwnien&e sample
================*================
 contrapositive
if p and q are two logical propo
sitions, then the contrapositive
of the proposition (p IMPLIES
q) is the proposition «NOT q)
IMPLIES (NOT p) ). The
contrapositive is logically
equivalent to the original
proposition.
_ control for a variable
to control for a variable is to try
to separate its effect from the
treatment effect, so it will not
confound with the treatment.
There are many methods that
try to control for variables.
Some are based on matching
individuals between treatment
and control; others use assump
tions about the nature of the
effects of the variables to ~ to
model the effect mathemati
cally, for example, using regres
sl0n.
 control group
the subjects in a controlled ex
periment who do not receive
the treatment.
control
there are at least three senses
of "control" in statistics: a
member of the control group,
to whom no treatment is given;
~ a controlled experiment, and to
; control for a possible confound
: ing variable.
I
:  controlled experiment
~ an experiment that uses the
; method of comparison to evalu
: ate the effect of a treatment by
~ comparing treated subjects
I with a control group, who do
; not receive the treatment.
:  controlled, randomised
I •
: expenment
I a controlled experiment in
~ which the assignment of sub
: jects to the treatment group or
~ control group is done at ran
; dom, for example, by tossing a
: com.
I
:  convenience sample
~ a sample drawn because of its
; convenience; not a probability
: sample. For example, I might
~ take a sample of opinions in
; Delhi (where I live) by just ask
: ing my 10 nearest neighbors.
I
: That would be a sample of con
I venience, and would be unlikely
: to be representative of all of
~ Delhi. Samples of convenience
~ are not typically representative,
; and it is not typically possible
: to quantify how unrepresenta
I
MMIIenIaeics======____.11
38 .. II.
tive results based on samples of 2. if P and q are two logical
convenience will be. i propositions, then the converse
T_. : of the proposition (p IMPLIES
q) is the proposition (q IM
_of __ of _..... .:.. i i PLIES p).
v.,..., 1> ..
I
. c.: ..... ."" ..... ! iconversion factor
a. 1.8UJIa!i1C9 COl •
> .. ,,' .,.,  'Ii>! 00. : relationship between two umts
. ':" :::  :: from different systems of mea
_001 i surement used to convert from
_ converge convergence ; . one system the other (e.g.,
a sequence numbers xl, x2, : centimeters corresponds to
x3 . . . converges if there is a 1 mch).
number x such that for any num i_convex polygon
ber E>O, there is a number k : a polygon having no diagonal
(which can depend on E) such lying outside the polygon.
that Ixj xl < E whenever j > k. I
. . ..  convex set
If such a number x eXISts, It 18 0 f 0 0 hich all
called the limit of the sequence a set 0 pomts om w 0 fSth
eg

1 x2 3 i ments connectmg pomts 0 e
x, , x 0 0 • 0 : set lie entirely in the set; There
 convergence in probability are three things one can do to
a sequence of random variables i see if a figure is convex look for
Xl, X2, X3 0 0 0 converges in : "dents", extend the segments
probability if there is a random (they shouldn't enter the fig
variable X such that for any jure), and connect any two
number E>O, the sequence of : points within the figure with a
numbers P( IXI XI < e), P( 1X2 segment (if any part of the seg
XI < e), P( IX3 XI < e), . 0 0 I ment lies outside the figure, it's
converges to 100%. : concave).
I
 converse : _ coordinate
1. (of a conditional statement), a number that identifies (or
the statement formed by ex I helps to identify) a point on a
changing the "if" and "then" : number line (or on a plane, or
f
"if.th" I .
parts 0 an  en statement : m space)
I
 coordinate geometry
the study of geometrically rep
resenting ordered pairs of num
bers
 coordinate plane
39
~ vector may be thought of as
; coordinates in a geometric
: sense.
I
:  coordinatised. line
~ a line on which every point is
; identified with exactly 1 num
: her and vice versa; a onedi
~ mensional graph. The distance
I between 2 parts on a
: coordinatised line is the abso
~ lute value of the difference of
I their coordinates.
I _ coplanar
~ lying in the same plane.
;  coprime
a plane in which every point is
identified with exactly 1 num
ber and vice versa; a twodi
mensional graph
: integers m and n are coprime if
_ coordinate proof ~ gcd(m,n) = 1.
a proof using coordinate geom ~ _ corollary to a theorem
etry. ; a theorem that is easily proved
_ coordinate system : from the first
set of ordered pairs used to 10 I
cate an object or point on the I
twodimensional plane.
 coordinates
a set of parameters that locate
a point in some geometrical I
space. Cartesian coordinates,
for instance, locate a point on a
plane or in physical space by I
projection onto perpendicular
I I .
p
lines through one single point, :  corre abon
the origin. The elements of any I relation between two or more
; variables. Frequently the word
M/lthemR,riu======== II
I • correlation
is used for Pearson's producr
moment correlation which is
the covariance divided by the
product of the standard
deviations,rxy=Sxy / Sx.Sy This
correlation coefficient IS + 1 or
1 when all values fall on a
straight line, not parallel to ei
ther axis. However, there are
also Kendall, Spearman,
tetrachoric, etc. correlations I
which measure other aspects of
die relation between two vari
ables. ,
• correlation coefficient
the correlation coefficif!nt r is
I a measure of linear association
between two (ordered) lists.
Two variables can be strongly
; correlated without having any
causal relationship, and two
variables can have a causal re
I lations hip and yet be
uncorrelated.
• corresponding angles
if two parallel lines are cut by a
transversal, corresponding
angles are translations of each
other along the transversal .
• cosine
I (of an acute angle) The ratio of
a measure of how nearly a the length of the adjacent side
scatterplot falls on a straight I to the length of the hypotenuse
line. The correlation coeffi I
in any right triangle containing
cient is always between 1 and I the angle.
+ 1. To compute the correla .
tion coefficient of a list of ; • coterminal angles
pairs of measurements (X,Y), : two angles that have the same
first transform X and Y indi terminal side
vidually into standard units. "I • countable set
Multiply correspopding ele ; a set is countable if its elements
ments of the transformed : can be,put in corre
pairs to get a single list of spondence with a subset of the
numbers. The correlation co ; integers. For eXaniple the sets
efficient is the mean of. that 7 3} {red, blue}
list of products. This page { , , '2 :1 0 1 2 }'
. . ... " "" ... ,
contams a lets y?U I {straiWtt, curly}, and the set of
generate" bIVarIate data WIth : all fractions are countable. If a
I
· ffi . I '
any corre anon coe ICIent YQU . set is not countable it is un
. ,
want.
41
II counterclockwise I criticalveUue
countable. The set of all real
numbers is uncountable.
• counterclockwise
in orientation, the direction in
which points are named when
· '
if travelling on the line, the in
terior of the figure is on the left
side.
• counterexample
given transformation has only
lone direction of covariants, but
: a full plane (four landmarks) or
hyperplane (five or more land
I marks) of invariants.
C<:MI:I'iAntl"lft.lrlcTII'rtSf.r ... .rflcllll'4 .. ,
(aC'lt .. rl'(HTl
A
a situation in a conditional for I
which the antecedent is true, but
the conditional is false; aka con I
tradiction
• counting techniques
a variety of methods used to
determine the total possible
outcomes, typically in a prob
ability situation, including the
multiplication principle, trees
and lists.
• covariant
a covariant of a particular shape
change is a shape variable
whose gradient vector as a func
tion of changes in any complete
set of shape coordinates lies
precisely along the change in
question. For transformations
of triangles, the relation be
tween invariants and covariants
is a rotation by 90 degrees in
the shapecoordinate plane. For
more than three landmarks a ,
• cover
; a confidence interval is said to
: cover if the interval contains the
true value ofl the parameter.
I Before the data are collected,
: the chance that the confidence
interval will contain the param
I eter value is the coverage prob
ability, which equals the confi
: dence level after the data are
I collected and the confidence in
terval is actually computed.
; • coverage probability
: the coverage probability of a
procedure for making confi
; dence intervals is the chance
: that the procedure produces an
interval that covers the truth.
• critical value
; the critical value in an hypoth
: esis test is the value of the test
I
Msthernades,=======II
42 crossseaional study I Ciunulati; Probability Distribution Pmu:tUm... II
statistic beyond which we has some or all of its digits re
would reject the null hypoth ; placed by letters or symbols and
esis. The critical value is set so : where the restoration of the
that the probability that the test original digits is required. Each
statistic is beyond the critical ; letter represents a unique digit.
value is at most equal to the sig • cube
nificance level if the null hy : a solid figure bounded by 6 con
pothesis be true. I
: gruent squares.
• crosssectional study I ub' 0'
. • • C IC equa on
a crosssectlonal study com . I 'al ti f d
pares different individuals to a po r
oml
equa on 0 e
each other at the same time ; gree .
it looks at a crosssection of a ;
population. The differences be
tween those individuals can cQn
found with the effect being ex ;
plored. For example, in trying :
to determine the effect of age
on sexual promiscuity, a cross ;
sectional study would be likely :
to confound the effect of age
with the effect of the mores the ;
subjects were taught as children:: C u1 u' 'D. b bili'ty
th Id
. di 'duals b I. um a ve .ero a
eo er m VI were pro . D' m'b u' F 0'
bl
. ed·th ...1:.a:' IS U on unc on
a y WI a very . (CDF)
towards proIDlS
b
. cu ; the cumulative distribution
Ity e younger su .. fun' f d . bl
Th
· uld be . rud ; etlon 0 a ran om varia e
us It wo Imp ent to . th h th th d
'b ...1:.a: • .: IS e c ance at e ran om
attn ute wuerences m proIDlS I . bl . 1 than ual
. th' C f. varia e IS ess or eq to
CUlty to e agmg process. ..: fun' f In bols
1 . rudinal tud I X, as a etlon 0 x, sym ,
ong! s y. : if F is the cdf of the random
• cryptarithm variable X, then F(x) = P( X < =
a number puzzle in which an ; x). The cumulative distribution
indicated arithmetical operation : function must tend to zero as x
I
II :....... .... 1 *=========!!!!!!4!!!!!!3
approaches minus inftnity, and  cyclic polygon . .
must tend to unity as x ap ; a polygon whose vernces lie on
proaches inftnity. It is a positive a circle.
function, and increases mono : _ cyclic quadrilateral .
tonically: ify > x, thenF(y) >= a quadrilateral that can be m
F(x). The cumulative distribu ; scribed in a circle.
tion function completely
characterises the probability dis ;  cylinder .
tribution of a random variable. : a solid whose surface COnsISts of
. all points on two in
 curved I parallel planes, along WIth pomts
a space wlth coordinates and : in their interiors (the bases of the
a distance such that cylinder), and all points line
the area of Circles, volume of I segments joining the two arcles.
spheres, etc. are not : The axis of the cylinder is the line
tional to the ap.propnate segment that joins the centers of
power of the radIUs, e. g., I the bases. The radius of the cyl
Kendall's shape space. In : inder is the radius of a base. An
curved spaces, the usual iI?tui  altitude of a cylinder is a line seg
tions about what "straight I ment between. and perpendicu
lines" can be expected to do : lar to the planes of the bases. The
. I
will be For mstaI?ce, : height of a cylinder is
corresponding to every trlan of an altitude. If the axIS IS per
gular shape i? Kendall's ; pendicular to the basc:s, then the
space, there IS. another t?at I! : cylinder is a right cylinder; oth
"as far fronl as 1?osslble, erwise, it's oblique.
just like there IS a pomt on the I •• •
surface of the earth as far as :  cylindric s?lid
possible from where you now the set pomts a re
sit. ; gion
 cycle
(of a periodic curve) One
tion of a curve that, when laid
out repeatedly endtoend,
forms the entire curve.
: including llle region an Its .lffi
age
;  cylindric surface
: the union of the bases and the
I
: lateral surface
II
44 =================* dartI dejieientcotmlintJte II
.dart
a concave kite.
are 52! (52 factorial)
possible orderings.
data . ; • decompose numbers
used as a basis for ; break up numbers into addends
reasonmg. . or factors.
• decagon • deductive reasoning
a tensided polygon reasoning accepted as logical
• decimal number ; from agreedupon assumptions
a number written to the base 10. : and proven facts.
I
• decision rule :  deductive system
a rule for comparing the out an arrangement of premises
come value of alpha with a ; and theorems, in which each
nominal alpha criterion level: theorem can be proved by de
(such as 0.05). An outcome reasoning using only the
value smaller (more extreme) ; premIses theo
than the nominal alpha criterion : rems, and m which each defini
level leads to a decision of sta tion uses only terms that have
tistical significance of the find ; been defmed previously in the
ing that the test statistic has a : system.
value other than its (null) • deficient coordinate
hypothesised value. in addition to landmark loca
• deck of cards ; tions, a digitiser can be used to
a standard deck of playing cards : supply information of other
contains 52 cards, 13 each of sorts. For example, a point can
four suits: spades, hearts, dia ?e used .to encode part of the
monds, and clubs. The thirteen ; about a curving arc
cards of each suit are {ace 2 3 : by Identifying the spot at which
4? 5,6,7,8,9, 10, jack, the lies farthest from some
king}. The face cards are {jack, I other unage structure (perhaps
queen, king}. It is typically as : another such curving arc). The
sumed that if a deck of cards is null model of independent
shuflled well, it is equally likely ; noise does not apply
to be in each possible ordering. ; to posItion along the tangent
45·
II MJkimtnumbwl
direction of the curve that is missing observation can be
digitised in this that ; computed as.
Cartesian coordinate IS defi  ; _ denominator
cient." The usual model of in : in the fraction xly, x is called the
dependent is numerator and y is called the
inapplicable m pnnclple for ; denominator.
such points.
 deficient number
a positive integer that is
than the sum of its proper diVl
sors.
 definition
a statement that clarifies or ex
plains the meaning of a word
or phrase.
;  dense line
: the line that contains the shon
est path between two points
 density
; the ratio of the mass of an ob
: ject to its volume.
I
:  density scale
the vertical axis of a histogram
; has units of percent per unit of
 degree : the horizontal axis. This is called
a unit of m,easure for angles and a density scale; it measures how
arcs equivalent to 1/360 of ro I "dense" the observations are in
tation around a circle. A right each bin ..
angle measures 90 degrees, . de . bl
s bolized b 90°. ;  depen nt vana e .
ym y : in regression, the vanable
 degrees of freedom . whose 'values are supposed to
given a set of parameters ; be explained by in the
mated from the data, the de : other variable (the the mdepen
grees of freedom" of s?me sta dent or explanatory variable).
tistic is the of ; Usually one regresses de
dent to : pendent variable on the mde
compute the stattsttc. For ex pendent variable.
ample the variance has n1 de ;
grees freedom because only :  derived
n1 of the observations are a. that IS a com
needed for its computation ; bmatton of two other measure
given the sample mean. The :
I
MMhemsnes=========== II
46
=========. describe I dijfermtUJte II
ments such as speed in miles per • diagonal
hour . lin
. I a e segment connecting two
• describe : nonconsecutive vertices of a
I
to explain orally or in writing. : polygon or polyhedron.
I di
• determine .• ameter
to know, or make it possible the. longest of figure. In
to know, all the characteristics ; a CIrcle, a diameter IS a chord
of a fi?ure. For example, : that passes through the center
three Sides determine a tri _ . of the circle.
angle; three angles do not de • diameter of a circle (or
termine a triangle. ; sphere)
• develop : the segment whose endpoints
to be in reasoning, ex are points <;>n a circle (or sphere)
ploratlon, conjecturing, using I the. center of the
manipulatives or sketches to gain : CIrcle as Its mldpomt; the length
understanding of concepts or re of that segment
lationships. I • difference of means
• develop fluency ; a test statistic of intuitive appeal
to. become skillful in working ; for. measuring difference in 10
With numbers both in accu . cation between two samples
racy and speed. data. Em
. ploymg this test statistic in an
• deviation I
: exact test defmes the pitman
a deviation is the difference
b d
I permutation tests(l or 2).
etween a atum and some :
reference value, typically the • differential calculus
mean of the data. In comput ; th.at part of calculus that deals
ing the SD, one fmds the rms : the opeation of differen
of the deviations from the tIatIon of functions.
mean, the differences be ; • differentiate
tween the individual data and ; to distinguish from other mem
the mean of the data. : bers of a class, based upon
some chosen properties or cri
; teria.
II tIi,gimetie I discrete line
47
• digimetic ~  direction
a cryptarithm in which digits ; the way a number goes positive
represent other digits. : or negative
I
• digit :  direction of a translation
in the decimal system, one of ~ the compass direction in which
the numbers 0, I, 2, 3, 4, 5, 6, ; a translation goes.
7, 8, 9. ~ _ disc
 dihedral angle : a circle together with its inte
the angle formed by two planes ~ rior.
meeting in space.
~  dis.crete distribution
c
 dilation
; a probability distribution of
: some statistic, based upon an
~ algebraic formula or upon re
I randomisation or upon actual
~ data, in which the cumulative
: probability increases in non
~ infmitesmal steps correspond
a nonrigid transformation that ; ing to noninfmitesmal weight
enlarges or reduces a geomet : associated with possible values
ric figure by a scale factor rela ~ of the statistic in question. This
tive to a point (the center of the ; situation is characteristic of
dilation). : randomisation distributions,
• dimensions ~ and also of test statistics which
the width, length, and height of ; are essentially discrete.
a plane or space figure ~ _ discrete line /,
_ diophantine equation : a line made of dots with space
an equation that is to be solved ~ in between their centers.
in integers.
• direct variation
a variation of the type y = kx;
the graph is a straight line
through the origin.
E
48
 discrete variable a score which can be used to
a qUantitative variable whose set i classify the specimen as belong
of possible values is countable. ing to one or another predefined
Typical examples of discrete group.
variables are variables whose i di·· tuall I
possible values are a subset of :  . sJOlnt or mu Y exc u
. . I sive events
the mtegers, such as SOCial Se . dis· .
. be th . two events are or mu
cunty num rs, e number of I tuall . I . ifth
I
· d d . yexc USlve e occurrence
peop e m a aJ.IlllY, ages roun e . f .. ·bI·th th
th
. I 0 one IS mcompan e WI e
to e nearest year, etc. D15crete f th th th .
. bl " hunky." C f occurrence 0 eo er; at 15,
vana es are c .. . con I ifth ' both ha
. . bl A di ey can t ppen at once
ttnuous vana e. screte ran I (ifth ha .
dom variable is one whose set ey) nal° outltcome m com
f ·bl al' bi I mon . wv en y, two events
A
o
POSdSI e v .ubes
l
eif· : are disjoint if their intersection
ran om vana e 15 screte I· th
d nl if
· ula . b IS e empty set.
an 0 y Its cum nve pro 
• displacement
the volume of fluid that rises
above the original fluid line
ability distribution function is a
stairstep function; i.e., if it is .
piecewise constant and only in
creases by jumps.
when a solid object is placed
i into the fluid .
• discriminant analysis .
a broad class of methods con idissection
cemed with the development of : the result of dividing a figure
rules for assigning unclassified into pieces.
objects/specimens to previously I • distance
defmed groups. ; 1. the distance between points
_ disctiminant function : A and B is written as AB
a discriminant function is used 2. this term has several mean
to assign an.observation to one I ings in morphometrics; it
of a set of groups. Linear dis : should never be used without a
criminant functions take a vec prefixed adjective to qualify it,
tor of observations from a e.g., Euclidean distance,
specimen and multiplies it by a i Mahalanobis distance,
vector of coefficients to produce
Procrustes distance, taxonomic
distance.
3. (of a translation), the length
of the translation vector be
tween a figure and its image.
 distance between 2 parallel
lines
the length of a perpendicular
segment between them
tion is zero for small enough
; (negative) values of x, and is
: unity for large enough values
of x. It increases monotoni
; cally: if y > x, the empirical
: distribution function evalu
ated at y is at least as large
I as the empirical distribution
; function evaluated at x.
:  distribution
the of a set of nu
I merical data is how their val
; ues are distributed over the
: real numbers. It is completely
characterised by the empiri
AF02.165_ ; cal distribution function.
: Similarly; the probability dis
tribution of a random vari
; able is completely
 distribution function,
: characterised by its probabil
empirical ity distribution function.
the empirical (cumulative) ; Sometimes the word "distri
distribution function of a set
. . : bution" is used as a synonym
of numencal data IS, for each I ti th .. al di 'b .
al I f h fi
. f' or e empirIc stn utIon
re va ue 0 x, t e ractIon 0 . f' h b b T
observations that are less or fit e pro a I tty
h I A I f
; Istn utIon unctIon.
t an or equa to x. p ot 0
the empirical distribution  distributive law
function is an uneven set of : the formula a(x+y)=ax+ay.
The of the _ distributive property
IS the adJa thepropertywhichrelatesmul
cent data, the height of the . tiplication and addition' the
stairs depends on how many formula, a(b + c) = a x b + a
data have same I x c.
value. The distnbutlon func :
50
~ ~ ~ ~ ~ ~ ~ ~ = *
. diPitlend I drawing II
• dividend ~ • domino
in the expression "a divided by ; two congruent squares joined
b", a is the divident and b is the : along an edge.
di
. I
VIsor. : • dot
• divisor ~ a description of a point in which
in the expression "a divided by ; the point has a definite size
b", a is the divident and b is the : • d bl I" h
di . I ou e me grap s
VIsor. : graphs in which two sets of data
• dodecagon. ~ are graphed at the same time,
a twelvesided polygon I connecting each set with line
• dodecahedron ; segments.
a solid figure with 12 faces. A : • doubleblind, double
regular dodecahedron is a ~ blind experiment
regular polyhedron with 12 I in a doubleblind experiment,
faces. Each face is a regular; neither the subjects nor the
pentagon. : people evaluating the subjects
I knows who is in the treatment
; group and who is in the control
: group. This mitigates the pla
~ ceho effect and guards against
; conscious and unconscious
: prejudice for or against the
~ treatment on the part of the
; evaluators.
; .draw
: to create a figure using num
d
" ~ bered scales on tools such as .
• omam . r u l ~
th d
. f fun' fi()' I ers and protractors.
e omamo a mon x IS •
the set of x values for which the ; • drawing
function is defined. ; a freehand picture using any
: tool.
II dual I eigenshapes
dual
(of a tessellation), the new
sellation formed by connectmg
(with line segments) the centers
of polygons with a cOI?mon
edge in another tessellanon.
 duodecimal number
system .
the system of numeration with
base 12.
 ecological correlation
the correlation between aver
ages of groups of
instead of individuals. EcologI
cal correlation can be mislead
ing about the association of in
dividuals.
51
 edge
; (of a solid) The intersection of
: two faces.
I
:  edgel
I an extension of the notion of
landmark to include partial in
: formation about a curve
through the landmark. An
; edgel specifies rotation of a di
: rection through a landmark,
extension along a direction
; through a landmark, or both.
: The formula for thinplate
splines on landmarks can be
I extended to encompass data
: about edgels as well. They are
I , .
: intended eventually to circum
vent any need for deficient co
; ordinates in multivariate mor
: phometric analysis.
I
:  Egyptian fraction
a number of the form l/x where
; x is an integer is called an Egyp
: tian fraction.
I
:  eigenshapes
principal components for out
; line data. An eigenshape
: analysis begins with the selec
tion of a distance function be
; tween pairs of outlines. At the
: end one gets "eigenshapes,"
I . f
: which have the propernes 0
principal component vectors
5
7
2 :enPeCtors I empirietdlmPof.."tII'lIJIes II
(uncorrelated, describing the· • ellipse
sample in decreasing order of ; a plane figure whose equation
variance) and also are outline: is x2ja
2
+y2jb2=1.
shapes themselves, so .that the • ellipsoid
scores for each of a solid figure whose equation is
the sample can be ; x2ja2+y2jb2+z2jc2=1.
to produce a new outhne :
shape that approximates it in I • elliptic Fourier analysis
some possibly useful way. : a type of outline analysis in
Eigenshapes apply to curves which differences in x and y
as relative warps apply to (and z) coordinates of
landmark shape. . an outline are fit separately as
a function of arc length by Fou
• eigenvectors I rier analysis.
in the equation given to define
eigenvalues, E contains the • elliptic geometry
eigenvectors. In the common ; a geometry in which there are
data analysis case, E is an or : no parallel lines.
thonormal. matrix (i. e., • empirical law of averages
EtE=I and .. When the Empirical Law of Aver
sorted by descen?IDg  ; ages lies at the base of the fre
ues, first IS : quency theory of probability.
that bnear combmatl0n of I This law: which is in fact an
that has the .greatest : assu'llp;ion how'the
vanance. The second elgenvec world works rather than a
tor is linear combination I mathematical 'or physical law,
of vanables that has the I states that if one repeats a ran
greatest variance of such com dom experiment over and
binations orthogonal to tl}e lover, independently and under
first, and so on. I "identical" conditions, the
• elementary function fraction of trials that result in
one of the functions: rational a given outcome converges to
functions, trigonometric func
tions, exponential functions, and
logarithmic functions.
I a limit as the number of trials
grows without bound.
II empty set I equidisttmt
53
• empty set • enumerable set
the empty set, denoted {} or 0, ; a countable set.
is the set that has no members.
; • epicyc10id
• endpoint convention
: the locus or path of a point on a
circle as it rolls around another
; circle.
in plotting a histogram, one
"must decide whether to in
clude a datum that lies at a
class boundary with the class
interval to the left or the right
of the boundary. The rule for
making this assignment is
called an endpoint convention. I
The two standard endpoint
conventions are to include the
left endpoint of all inter I
vals and exclude the right, ex
cept for the rightmost class
interval, which includes both I
Eptcycloid
Coustle Cuns (wrt horizontal reus)
of its endpoints, and to in • equianglular
clude the right endpoint of all : having angles of the same mea
class intervals and exclude the I sure
left, the leftmost in • equiangular polygon
mcludes both of ; a polygon all of whose interior
ItS endpomts. : angles are equal.
• . equichordal point
the pomts at the ends of a line a point inside a closed convex
segment or arc. ; curve in the plane is called an
• ends of a kite : equichordal point if all chords
the common vertices of the through that point have the
equilateral sides of a kite ; same length.
.enneagon
a ninesided polygon
; • equidistant
: equally distant.
I
54
• equilateral
equal in length
• equilateral polygon
a polygon all of whose sides are
equal.
• equilateral triangle
elJuilateml I esti'flUl,tqr II
I in place of a test upon the de
I scriptively more valid one, with
corresponding savings in
I amount of computation re
I quired. An example of such
a triangle whose sides are equal I
in length
equivalent test statistics occurs
for the situation of comparison
of levels of a single interval
scale variable between two
groups. In this situation, the
I descriptively valid statistic, as
defined for the Pitman permu
tation test, is the difference of
I means, but simpler equivalent
test statistics include the mean
for one designated group, or
I (most simply) the total of scores
I in one designated group.
I • error tolerance
• equivalent test statistic the value allowable above and
within a randomisation set, it is I
possible that two different sta below a number or its approxi
. I mation.
tistics may be interrelated in a
manner which is provably I • escribed circle
monotonic irrespective of the an escribed circle of a triangle
data. In such a situation a is a circle tangent to one side of
randomisation test performed I the triangle and to the exten
on either of these test statistics sions of the other sides.
will necessarily have the same : • estimatot:
outcome in terms of alpha. If I an estimator is a rule for "guess
one of the statistics is of good ; ing" the value of a population
descriptive validity whereas the I parameter based on a random
other is simpler to compute, : sample from the population. An
then a raridomisation test upon estimator is a random variable,
the simpler statistic may be used . because its value depends on
II
II euclitleandistancema;t'l'ixanalysis I ;"",t==========5=5
which particular sample is ob ~  Euler segment
tained, which is random. A ca ; the line segment containing the
nonical example of an estimator : centroid of a triangle, whose
is the sample mean, which is an ~ endpoints are the orthocenter
estimator of the population ; and the circumcenter of the tri
mean. : angle.
I
 euclidean distance matrix :  Euler's formula for poly·
analysis ~ hedrons octahedron
EDMA. A method for the sta I an eightsided polyhedron. The
tistical analysis of full matrices of : regular octahedron is one of the
all interlandmark distances, av ~ Platonic solids.
eraging elementwise within ~ _ Euler's constant
samples, and then comparing I the limit of the series 1/1 + 1/
those averages between samples ~ 2+1/3+ ... +1/nln n as n goes
by computing the ratios of cor : to infInity. Its value is approxi
responding mean distances. ~ mately 0.577216 .
• euclidean space
a space where distances be
tween two points are defmed as
Euclidean distances in some
system of coordinates.
• euler line
the line through a triangle'S cir
cumcenter, orthocenter, and
centroid. Named after Swiss
I _ even function
; a function f(x) is called an even
: function if f(x) =f( x) for all x.
I
:  even node
~ a node that has an even num
; ber of arcs
;  even nwnber
: an integer that is divisible by 2.
I
mathematician and physicist : _ event
Leonhard Euler. ~ an event is a subset of outcome
A Euler or nine.point circIc of ABC. ; space. An event determined by
: a random variable is an event
~ of the form A=(X is in A).
; When the random variable X is
: observed, that determines
~ whether or not A occurs: if the
56
~ ~ ~ ~ ~ ~ ~ = .
value of X happens to be in A,
A occurs; if not, A does not oc
cur.
• exact binomial test
a statistical test referring to the
binomial distribution in its ex
~ group for the development and
; promulgation of the ideas of re
: randomisation statistics.
I
: .excenter
~ the center of an excircle.
I • excircle
act algebraic form, rather than
through continuous approxima ~ an escribed circle of a triangle.
tions which are used especially I
where sample sizes are substan :
tial.
• exact number
a numerical result that has not I
been rounded or estimated.
• exact test
~ • exclusive OR
~ one or the other, but not both
 the characteristic of a re
randomisation test based upon
exhaustive rerandomisation,
that the value of alpha will be
fixed irrespective of any random I • exhaustive re
sampling of randomisations or ~ randomisation
uponanydistributionalassump : a series of samples from a
tions. Notable examples are the ~ randomisation set which is
exact binomial test, fisher test, ; known to generate every
the Pitman permutation tests ( 1 randomisation. In particular,
and 2), and various nonpara I sampling which generates every
metric tests based upon ranked I randomisation exactly once.
data.
; • exhaustive
• exactstats : a collection of events {AI, A2,
this is the name of the academic ~ A3, ... } is exhaustive if at least
initiative which produced this ; one of them must occur; that
present glossary. exactstats is : is, if S = Al U A2 U A3 U ...
a closed email based discussion ~ where S is the outcome space.
II existentUdstll:tement I e:cpectedi = nxpi
57
• existential statement
a conditional that uses the word
'same'
• expansion
a size change where k is greater
than I
pected value of a constant a
; times a random variable X is the
constant times the expected
: value of X (E(axX ) = aX
I E(X».
t • expected i = nxpi
• expectation, expected if the model be correct, we
value : would expect the n trials to re
the expected value of a random in o.utcome i about nxpi
is the longterm limit ; tunes, gIve or take a bit. Let
mg average of its values in in : observed i denote the number
If'
dependent repeated experi : 0 an outcome of type i
ments. The expected value of t occurs m the n trials, for i = I,
the random variable X is de : 2, ... , k. The chisquared sta
t ••
noted EX or E(X). For a discrete : summarises the discrep
random variable (one that has I anCles between the expected
a countable number of possible : number of times each outcome
values) the expected value is the occurs (assuming that the
weighted average of its possible model is and the observed
where the weight as ; number of tImes.each outcome
sIgned to each possible value is : occurs, by summmg the squares
the chance that the random vari of discrepancies,
able takes that value. One can ; normalIsed by the expected
think of the expected value of a : over all the catego
random variable as the point at nes: chisquared = (observed1
which its probability histogram ; expected1)2/expected1 + (ob
would balance, if it were cut out : served2 expected2)2/expected2
of a uniform material. Taking + . . . + (observed k expected
the expected value is a linear t As the sample
operation: if X and Y are two : SIZe n mcreases, if the model is
random variables, the expected c.OITecr, the distribu
value of their sum is the sum of t of the chisquared statistic
their expected values (E(X + Y) : IS approximated increasingly
= E(X) + E(Y», and the ex well by the chisquared curve
MsthlJllllilties============ II
58
txperiment I ~ t m e n t /I
with (#categories 1) = k 1 de ~ _ explanatory variable
grees of freedom (d.f.), in the ; in regression, the explanatory
sense that the chance that the : or independent variable is the
chisquared statistic is in any ~ one that is supposed to "explain"
given range grows closer and; the other. For example, in ex
closer to the area under the Chi amining crop yield versus quan
Squared curve over the same tity of fertilizer applied, the
range. This page illustrates the I quantity of fertilizer would be
sampling distribution of the chi the explanatory or independent
square statistic. variable, and the crop yield
_ experiment I would be the dependent vari
what distinguishes an experi I able. In experiments, the ex
ment from an observational planatory variable is the one
I that is manipulated; the one
study is that in an experiment,
the experimenter decides who I
receives the treatment.
 experimental design
that is observed is the depen
dent variable.
 explicit form
I a formula for any term of a se
quence given the number of the
this term overtly refers to the
planning of a pJrocess of data I
collection. The term is also used I term.
to refer to the information nec  exploration  indirect
essary to describe the interre I measurement clinometer
lations hips within a set of data. I a tool for measuring angle of
Such a description involves con elevation or depression, con
siderations such as number of ~ sisting of an edge to sight along,
cases, sampling methods, iden I a plumb line, and a protractor.
I _ expoential function to
base a
tification of variables and their
scaletypes, identification of re
peated measures and replica I
tions. These considerations are I
the function f(x)=a
x
•
essential to guide the choice of :  exponent
TEST STATISTIC and the pro ~ in the expression xl', x is called
cess of RERANDOMI· the base and y is called the ex
SATION. ponent.
II '"P
unmtisl
funaiun I foetorial
• exponential function
the function f(x) =cX.
Future Value as an Exponential FlIDction
Amction
y=ed"
Relatillg to OUr c ....
FV : PV(l + If : PVoi"'J+'" = ed"
y=FV
a=PV
b=lt(1+1l, 1= ... ·1
• expression
combination of numerals or
numerals and variables that in
dicate a ftnite number of opera
tions, not an equation.
• exradius
an exradius of a triangle is the
radius of an escribed circle.
• exterior angle
1. (of a triangle), an angle that
59
~ • extrapolate
; to use given information to pre
: dict values beyond the set of
~ given values using either a for
; mula or a reasonable estimate.
; • face .
: (of a polyhedron) one of the
~ polygons and its interior form
~ ing the surface of a polyhedron.
I • face angle
: the plane angle formed by ad
~ jacent edges of a polygonal
~ angle in space.
I • factor (noun)
~ an exact divisor of a number.
: This 7 is a factor of 28.
I
: • factor (verb)
I to fmd the factors of a number.
forms a linear pair with one of ~ • factor analysis
the interior angles of a triangle. ; factor analysis is a multivariate
2. (of a polygon), an angle that : technique for describing a set of
forms a linear pair with one of ~ measured variables in terms of
the interior angles of a polygon. ; a set of causal or underlying
• exterior of an angle
the nonconvex set formed by an
angle that measures less than
180 degrees.
• externally tangent
(circles) intersecting at exactly
one point, with neither circle
inside the other.
: variables. A factor model can be
~ characterised in terms of path
; diagrams to show relations be
: tween measured variables and
~ factors.
~ • factorial
; the factorial operator is appli
: cable to a nonnegative integer
~ quantity. It is notated as the
60
foirbet IfibCl' II
postfixed symbol 'I'. The result neously (the nwnber of Type I
ing value is the product of the ; errors divided by the nwnber of
increasing integer values from: rejected null hypotheses), with
I up to the value of the argu the convention that if no hy
ment quantity. For instance : 3! ; pothesis is rejected, the false
is lx2x3 = 6. By convention O! : discovery rate is zero.
is taken as producing the value ..... __ :1: fr cti"
1 f: . al val . J.CUlllllar a ons
. h mcreas.e commonly used fractions such
rapt y Wlty 1 m . ; as halves, thirds, fourths, fifths,
va ue; t diS .raP
th
l : sixths, eighths and tenths.
gro IS represente 10 e
similarly rapid growth in num . • family tree
bers of combinations.. hierarchy; tower or pyramid of
• fair bet power or importance
a fair bet is one for which the I • farey sequence
expected value of the payoff is the sequence obtained by ar
zero, after accounting for the : ranging in nwnerical order all
cost of the bet. For example, the proper fractions having de
suppose I offer to pay you Rs.2 ; nominators not greater than a
if a fair coin lands heads, but : given integer.
you must ante up Rs.l to play. • Fermat number
Your expected payoff is  Rs.I + I a nwnber of the form (22n+ I)
Rs.OxP(tails) + :
Rs.2xP(heads) =  Rs.l + •
Rs.2x50% = Rs.O. This is a I a parabolic Sptral.
fair betin the long run, if you ; • fiber
this bet over and over : the set of preshapes (configu
agam, you would expect to rations that have been centered
break even. ; at the origin and scaled to unit
• false discovery rate : centroid size) that differ only by
in testing a collection ofhypoth a rotation. It is the path,
eses, the false discovery rate is I through preshape space, fol
the fraction of rejected null hy lowed by a centered and scaled
potheses that are rejected erro
II Fibonacci numb",. I finite element =======",,6=1
configuration under all possible • figure
rotations. ; a representation of an object by
• Fibonacci number : the coordinates of a specified set
a member of the sequence 0, 1, of points, the landmarks.
1, 2, 3, 5, ... where each num • figure space
ber is the sum of the previous ; the 2por 3pspace of figures, i.
two numbers. : e., the original coordinate data
T .... tIIe ..... ""'wm ...... _ ...........
mnabem 1ft the pd8m.
• field properties
closure for addition and mul
tiplication, commutative for
addition and multiplication,
associative for addition and
multiplication, identity for
addition and multiplication,
inverse for addition and mul
tiplication, distributive for
multiplication over addition.
• figurate numbers
polygonal numbers
I
. vectors.
• finite element scaling
I analysis
: without the word "scaling," fi
nite element analysis is a com
I putational system for con
tinuum mechanics that esti
: mates the deformation (fully
detailed changes of position of
; :ill component particles) that are
: expected to result from a speci
fied pattern of stresses (forces)
; upon a mechanical system. As
: applied in morphometrics,
FESA solves the inverse prob
; lem of estimating the strains
: representing the hypothetical
forces that deformed one speci
I men into another. These results
: are a function of the "finite ele
ments" into which the space
I between the landmarks is sub
; divided. FESA can be compared
: with the thinplate spline, which
interpolates a set of landmark
; coordinates under an entirely
: different set of assumptions.
I
62
~ ~ ~ ~ ~ ~ ~ ~ *
Jinitegroup I Fisher test II
• finite group
a group containing a finite num
ber of elements.
• finite population correc
tion
~ exactly once), and the SE
; should be zero. This is indeed
: what the finite population cor
~ rection gives (the numerator
; vanishes).
when sampling without replace ; • Fisher test
ment, as in a simple random : named after the statistician RA
sample, the SE of sample sums ~ Fisher. This is an exact test to
and sample means depends on I examine whether the pattern of
the fraction of the population: counts in a 2x2 cross classifica
that is in the sample: the ~ tion departs from expectations
greater the fraction, the smaller ~ based upon the marginal totals
the SE. Sampling with replace ; for the rows and columns. Such
ment is like sampling from an : a test is useful to examine dif
infinitely large population. The ~ ference in rate between two bi
adjustment to the SE for sam ; nomial outcomes. The
pling without replacement is : randomisation set consists of
called the ftnite population cor ~ those reassignments of the
rection. The SE for sampling ; units which produce tables with
without replacement is smaller : the same row and column to
than the SE for sampling with ~ tals as the outcome. The
replacement by the fmite popu ; randomisation set will thus con
Iation correction factor ((N n)/ : sist of a number of tables with
(N 1»'12. Note that for sample ~ different respective patterns of
size n=l, there is no difference I counts; each such table will have
between sampling with and a number of possible
without replacement; the fmite randomisations which may be
population correction is then I a very large number. For this
unity. If the sample size is the test there are several reasonable
entire population of N units, test statistics, including : the
there is no variability in the re ~ count in anyone of the 4 cells,
sult of sampling without re ; chisquared, or the number of
placement (every member of : randomisations for each 2x2
the population is in the sample ~ table with the given row and
II =======MJJthemlJnes
I
II Fisher'sl'&ll&ttest Iflexibly
63
column totals; these are equiva ~ pothcsis be true, the two
lent test statistics. The calcula ; s a m p l ~ s arc like one larger
tion for the Fisher test is rela : sample frum a single population
tively undemanding ~ of zeros and ones. The alloca
computationally, making refer ; tion of ones between the two
ence to the algebra of the hy : samples would be expected to
pergeometric distribution, and ~ be proportional to the relative
the test was widely used before , sizes of the samples, but would
the appearance of computers. : have some chance variability.
This test has historically been ~ Conditional on G and the two
regarded as superior to the use ~ sample sizes, under the null
of chisquared where sample ; hypothesis, the tickets in the
sizes are small. Statistical tables : first sample are like a random
have been published for the ~ sample of size ni without re
Fisher test for a number of ; placement from a collection of
small 2x2 tables defined in : N = ni + n2 units of which G
terms of row and column totals. ~ are labelled with ones. Thus,
_ Fisher's exact test ; under the null hypothesis, the
for the equality of two percent ~ number of tickets labeled with
.. ones in the first sample has
ages, Consider two populations
of zeros and ones. Let pI be the ' (conditional on G) an hyper
P
roportion of ones in the frrst : geometric distribution with pa
~ rameters N, G, and nl. Fisher's
population, and let p2 be the , exact test uses this distribution
proportion of ones in the sec
ond population. We would like ~ to set the ranges of observed
: values' of the number of ones in
to test the null hypothesis that
pI = p2 on the basis of a simple ~ the first sample for which we
random sample from each; would reject the null hypoth
population. Let ni be the size ~ eSlS.
of the sample from population :  flexibly
1, and let n2 be the size of the ~ usually applied to computation,
sample from population 2. Let ; where students should be able
G be the total number of ones : to mentally manipulate num
in both samples. If the null hy ~ bers and components of num
64
bers to create a solution to a ~ • foot of altitude
problem. i the intersection of an altitude of
• floor function a triangle with the base to which
I it is drawn.
the floor function of x is the
greatest integer in x, i.e. the ~ • foot of line
largest integer less than or i the point of intersection of a line
equal to x. : with a line or plane.
• flowchart • footballshaped
I
a concept map that shows a: scatterplot
stepbystep process. Boxes I in a footballshaped scanerplot,
represent the steps, and arrows most of the points lie within a
connect the boxes to order the : tilted oval, shaped moreorIess
process. I like a football. A football
• flowchart proof shaped scatterplot is one in
which the data are
a logical argument presented in
the form of a flowchart. I homoscedastically scattered
about a straight line.
• focal chord
a chord of a conic that 'passes i • form
through a focus. in morphometries, we repre
• focal radius
sent the form of an object by a
point in a space of form vari
a line segment from the focus ables, which are measurements
of a geometric object that are
c
tIDchanged by translations and
rotations. If you allow for re
flections, forms stand for all the
figures that have all the same
interlandmark distances. A
form is usually represented by
one of its figures at some speci
fied location and in some speci
of an ellipse to a point on the I fied orientation. When repre
perimeter of the ellipse. sented in this w a ~ ~ location and
II ==========Math_ti&s
II form space Ifranu. mmpli"9franu * ========6=5
orientation are said to have been
"removed."
_ form space
~ (structural complexity at
; smaller scales is mathematically
: indistiguishable from that at
~ largerscales) over all scales con
; sidered.
t h ~ space of figures with differ
ences due to location and ori
entation removed. It is of 2p3 I
dimensions for twodimen
sional coordinate data and 3p6
dimensions for threedimen I
sional coordinate data.
BON Size Offset
Width 540
2 2 2 Height 110
3 .. 3
.. 13 ..
~
5 22 5
C!!!D
6 33 6
_ formula 44 7 I Dd.ulll
a concise statement expressing
the symbolic relationship be I
tween two or more qu;mtities.
8
9
10
"
55
66
77
88
8
O!!D
9
.10
"
~
00
 Fourier analysis
in morphometries, the decom
position of an outline into a
weighted sum of sine and co
sine functions. The chapter by
Rohlf in the Blue Book pro
vides an overview of this and
other methods of analysing out
line data.
 fourier series
a periodic function with period
2 pi.
12 111 12
I
:  fraction
~ an expression of the form alb . .
~  frame, sampling frame
I a sampling frame is a collec
~ tion of units from which a
: sample. will be drawn. Ideally,
I the frame is identical to the
; population we want to learn
: about; more typicaHy, the
~ frame is only a subset of the
; population of interest. The
_ fractal : difference between the frame
a selfsimilar geometric figure. ~ and the population can be a
_ fractal dimension ; source of bias in sampling de
. sign, if the parameter of rn
a measure of the complexity of I terest has a different value for
a structure assuming a consis I the frame than it does for the
tent pattern of selfsimilarity
population. For example, one
MRthmuJti&s======= II
66
~ ~ ~ ~ ~ ~ ~ = *
frelJumcy I funetion rule II
might desire to estimate the ~ (fraction or percentage) of ob
current annual average in ; servations in different ranges,
come of 1998 graduates of : called class intervals.
the University of Delhi. I pro I
pose to use the sample mean
income of a sample of gradu
ates drawn at random. To fa I
cilitate taking the sample and I
contacting the graduates to
obtain income information I 1
from them, I might draw I J
names at random from the list
of 1998 graduates for whom
the alumni association has an I
accurate current address. The I
population is the collection of :


..


1998 graduates; the frame is I
those graduates who have cur  frustum
rent addresses on file with the for a given solid figure, a re
alumni association. If there is I lated figure formed by two par
a tendency for graduates with I allel planes meeting the given
higher incomes to have upto solid. In particular, for a cone
date addresses on file with the I or pyramid, a frustum is deter
alumni association, that would I mined by the plane of the base
introduce a positive bias into and a plane parallel to the base.
the annual average income es I NOTE: this word is frequently
timated from the sample by I incorrectly misspelled as
the sample mean. frustrum.
_ frequency
the number of times a value
occurs in some time interval.
_ frequency table
a table listing the frequency
(number) or relative frequency
_ function rule
I the set of operations that de
I scribes the process that takes
the independent variable and
transforms it into the dependent
I variable in a consistent way.
• fundamental region
a region used in a tesselation
• fundamental rule of
counting
67
I variance and correlation of the
I variables in measuring dis
tances between points, i. e., dif
I ferences in directions in which
if a sequence of experiments or
trials TI, T2, T3, ... , Tk could I
result, respectively, in nl, n2 n3,
I in which there is more varia
· .. , nk possible outcomes, and
I there is less variation within
groups are given greater weight
than are differences in directions
the numbers nl, n2 n3, ... ,nk I tion.
do not depend on which out
comes actually occurred, the
entire sequence of k experi I
p
ments has nl x n2 x n3 x
x nk possible outcomes.
• game theory
= mill IIx  eil
ceo}
• generalised superimposi
tion
a field of study that bridges I
mathematics, statistics, eco
nomics, and psychology. It is
used to study economic I
behaviour, and to model conflict
between nations, for example,
"nuclear stalemate" during the I
I the superimposition .ofa set of
configurations. onto their con
sensus configuration. The fit
I ring may involve leastsquares,
Cold War.
• gaussian curve
a normal curve.
• generalised distance
d. A ,synonym for Mahalanobis
distance. Defined by the equa
tion for two row vectors x. and
•
Xj for two individuals, and p
variables as: , where S is the pxp
variancecovariance matrix. It
takes into consideration the
resistantfit, or other algo
rithms and may be strictiy or
I thogonal or allow affine trans
I formations.
; • geoboard
a flat board into which nails
I have been driven in a regular
I rectangular pattern. These nails
represent the · lattice points in
the plane.
=68=' =D=eodeS'ie='=I=Beomem&==' "
• geodesic
the arc on a surface of shortest
length joining two given points.
The expected value of the geo
. metric distribution is lip, and
its SE is (lp)V2/p.
• geodesic distance • geometric mean
the length of the shortest path I the nonnegative number whose
between two points in a suitable , square is the product of two
geometric space (one for which given nonnegative numbers;
curving paths have lengths). On I the side of a square having the
a sphere, it is the distance be I same area as a rectangle whose
tween two points as measured length and width are given.
along a great circle. ,. ________________ .. __ ,
• geodesy
a oranch of mathematics deal
ing with the shape, size, and
curvature of the Earth.
:! 0 I i
. OJ
, ____ I
• geometric distribution ... 
." ...
the geometric distribution de I • geometric morphometrics
scribes the number of trials · up , geometric morphometrics is a
to and including the first suc collection of approaches for the
cess, in independent trials with I multivariate statistical analysis
the same probability of success. I of Cartesian coordinate data,
The geometric distribution de usually (but not always) limited
pends only on the single param to landmark point locations.
eter p, the probability of success I The "geometry" referred to by
in each trial. For example, the the word "geometric" is the
number of times one must toss geometry of Kendall's shape
a fair coin until the first time I space: the estimation of mean
the coin lands heads has a geo shapes and the description of
metric distribution with param sample variation of shape using
eter p = 50%. The geometric I the geometry of Procrustes dis
distribution assigns probability tance. The multivariate part of
px(l p)klto the event that it geometric morphometrics .is'
takes k trials to the first success. ' usually carried out in a linear
II
IIBeometrie probllhility 1 probllhility I B ~ ~ 1 symmetry .
69
tangent space to the nonEu
clidean shape space in the vicin
ity of the mean shape. More
generally, it is the class of mor
phometric methods that pre
serve complete information
about the relative spatial ar
rangements of the data
throughout an analysis. As such,
these methods allow for the
visualisation of group and indi
vidual differences, sample
variation, and other results in
the space of the original speci
mens.
 geometric probability 1
probability
~  geometry
; the branch of mathematics that
: deals with the nature of space
~ and the size, shape, and other .
; properties of figures as well as
: the transformations that pre
I th .
: serve ese propernes.
I •
:  gergonne pomt
I in a triangle, the lines from the
: vertices to the points of contact
~ of the opposite sides with the
I inscribed circle meet in a point
~ called the Gergonne point.
;  given
: information assumed to be true
~ in a proof
(of an event) Its likely outcome, ~  glide reflection
expressed as the ratio of the ; an isometry that is a composi
number of successful outcomes : tion of a translation (glide) and
of the event to the number of ~ a reflection over a line that is
possible outcomes. ; parallel to the translation vec
: tor.
 geometric progression I
a sequence in which the ratio of :
each term to the preceding ~
term is a given constant.
 geometric series I
a series in which the ratio of :
each term to the preceding I
term is a given constant. ;  glidereflectional symme
: try
 geometric solid ~ the property of a geometric
the bounding surface of a 3di
~ figure that it coincides with its
mensional portion of space.
70
image under some glide re I
flection.
 gnomon magic square
a 3 X 3 array in which the ele
ments in each 2 X 2 corner I
have the same sum.
 gold standard
the gold standard is the form _ golden spiral
of test which : a spiral through vertices of
t? the distnbu I nested golden rectangles
non, for a gIven test statistic : .
experimental design. This 
Involves exhaustive randomi I a graph IS Said to be graceful if
sation. Other randomisation : you can number the n vertices
may reasonably be with the integers from 1 to n
Judged by comparison with I and label each edge with
this form. : the dIfference between the
numbers at the vertices, in such
I a way that each edge receives a
: different label.
 goldback's conjecture
if n is an even number greater
than 2, then there are always
2 prime numbers whose sum
is n
I
: _ grad (0" grade)
1/100th df a right angle
I
_ golden ratio .  grade
the ratio of two numbers ; the tilt of a reallife object in
(larger number : smaller; relation to the horizontal, often
number) whose ratio to each : to determine how steep a
other equals the ratio of their hill IS
sum to the larger number. ;  graph
_ golden rectangle ; a is a set of points (called
a rectangle in which the ratio : vertices) and a set of lines
of the length to the width is edges) joinging these
the golden ratio. vernces.
IloraPhofaveraoes Igrowingpattem .. ==========7=1
_ graph of averages  greatest common factor
for bivariate data, a graph of ; the greatest number that is a
averages is a plot of the aver : factor of each of the given num
age values of one variable (say bers.
y) for small of values of _ greatest lower bound
the. other van able (say x), ; the greatest lower bound of a
the value of s.econd set of real numbers, is the larg
vanable (x) at the ffildpomts of I est real number that is smaller
the ranges. I than each of the numbers in the
 graph theory I set.
the mathematics of complicated : _ grid
networks I I ' f
: a tesse atlOn 0 congruent
, squares sometimes used to
measure distance
.  group '.
a mathematical system consist
ing of elements from a set G
; and a binary operation * such
: that
x*y is a member of G whenever
• great circle ; x and yare
a circle on a sphere with a di : (x*y) *z=x* (y*z) for all X, Y, and
ameter equal to that of the z
sphere. The shortest path con ; there is an identity element e
necting two points on the sur : such that e*x=x*e=e for all x
face of a sphere lies along the each member x in G has an in
great circle passing through the ; verse element y such that
points. ; x*y=y*x=e
• greatest common divisor :. growing pattern
the greatest common divisor of a pattern where the number
a sequence of integers, is the I of objects in the pattern in
largest integer that divides each creases from term to term.
of them exactly.
72
• growth formula
I • harmonic mean
either a linear · or exponential
equation that describes the
growth over time.
I the harmonic mean of two
numbers a and b is 2abj(a + b).
• growth pattern
a set of values usually visualised I
by plotting points on a grid and
fitting either a linear or an ex
pOnential equation to the scat I
tel'" plot.
• halfline
a ray.
• halfplane
Hannoni c Mean
1 n
HM==
1 n 1
L
n i = 1 Xi
1 1 n 1
HM = L
n i = 1 Xi
• hectare
n 1
L 
i = 1 Xi
the part of a plane that lies on I
one side of a given line.
I a unit of measurement in the
metric system equal to 10,000
square meters (approximately
2.47 acres).
• Hankel matrix
a matrix in which all the ele
ments are the same along any I
diagonal that slopes from north
east to southwest.
I • height
l.the length of an altirude.
• harmonic analysis 2. (of a prism), the length of an
altitude.
the study of the r.epresentation I
of functions by means of linear
I • helix
operations on characteristic sets the path followed by a point
of functions.
moving on the surface of a right
• harmonic division I circular cylinder that moves
a line segment is divided har I along the cylinder at a constant
monically by two points when ratio as it moves around the
it is divided externally and in I cylinder. The parameteric equa
ternally int he same ratio. I tion for a helix is
x=a cos t
y=a sin t
I z=bt
1/ hemisphere I ~ m m
_ hemisphere
half of a sphere including a
great circle as its base.
 heptagon
a polygon with 7 sides.
_ heronian triangle
a triangle with integer sides and
integer area.
 heteroscedasticity
"Mixed scatter." A scatterplot
or residual plot shows
heteroscedasticity if the scatter
in vertical slices through the
plot depends on where you take
the slice. Linear regression is
not usually a good idea · if the
data are heteroscedastic.
 hexagon
a polygon with 6 sides.
 hexagonal number
a number of the form n(2nl) .
 hexagonal prism
a prism with a hexagonal base.
73
. = = = = = = = = = = ~ = = = = ~
~  hexahedron
; a polyhedron having 6 faces.
: The cube is a regular hexahe
~ dron.
~  hexomino
; a sixsquare polyomino.
;  hidden lines
: broken lines used to signify
~ lines that normally wouldn't be
I seen in a drawing
;  hierarchy
: a chart that shows varying lev
~ els of importance
I h"
:  Istogram
I a histogram is a kind of plot that
~ summarises how data are dis
: tributed. Starting with a set of
~ class intervals, the histogram is
; a set of rectangles ("bins") sit
: ting on the horizontal axis. The
~ bases of the rectangles are the
; class intervals, and their heights
: are such that their areas are pro
~ portional to the fraction of ob
; servations in the corresponding
: class intervals. That is, the
~ height of a given rectangle is
I the fraction of observations in
: the corresponding class interval,
~ divided by the length of the cor
I responding class interval. A his
~ togram does not need a verti
: cal scale, because the total area
I
74
=========*
historical controls .' homology II
of the histogram must equal I the outcome other than the
100%. The units of the vertical I treatment tend to change over
axis are percent per unit of the time as well. (In this example,
horiwntal axis. This is called I the level of other potential car
the density scale. The horiwn I cinogens in the environment
tal axis of a histogram needs a also could have changed.)
 homeomorphism
a onetoone continuous trans
I formation that preserves open
and closed sets.
scale. If any observations coin
cide with the endpoints of class I
intervals, the endpoint conven
tion is important. This page
contains a histogram tool, with I
controls to highlight ranges of ~
values and read their areas.
 historical controls
sometimes, the a treatment I
group is compared with indi
viduals from another epoch
who did not receive the treat I
ment; for example, in studying I
the possible effect of fluori  homology
dated water on childhood can I the notion of homology bridges
cer, we might compare cancer I the language of geometric
rates in a community before morphometrics and the lan
and after fluorine was added to I guage of its biological or biom
the water supply. Those indi I athematical applications. In
viduals who were children be theoretical biology, only the ex
fore fluoridation started would I plicit entities of evolution or
comprise an historical control I development, such as mol
group. Experiments and stud ecules, organs or tissues, can be
ies with historical controls tend "homologous." Following
to be more susceptible to con I D' Arcy Thompson, morpho
founding than .those with con metricians often apply the con
temporary controls, beca.use cept instead to discrete geomet
many factors that might affect I ric structures, such as points or
\I htmumwrpmsm I hypergeometric
curves, and, by a further exten in vertical slices through the
sion, to the multivariate de ; plot does not depend much on
scriptors (e.g., partial warp : where you take the slice. C.f.
scores) that arise as part of heteroscedasticity.
most multivariate analyses. In • horizontal line
this context, the term "homolo ; a line whose slope is zero
gous" has no meaning other
than that the same name is used ; • hyperbola
for corresponding parts in dif : a curve with equation x
2
/a
2
_y2/
•. b
2
=1.
ferent species or developmen
tal stages. To declare something • hyperbolic spiral
"homologous" is simply to as • the curve whose equation in
sert that we want to talk about : polar coordinates is r*theta=a.
processes affecting such struc • .
tures as if they had a consistent : • hyperboloid '
biological or biomechanical a geometric solid whose equa
. S"l 1 d I • tion is x
2
/a
2
+ y2/b
2
_Z
2
/C
2
= 1 0[,
mearung. 1m1 ar y, to ec are
. l' ( h thin.: x
2
/a
2
+ y2/b
2
_z , 2/C,2=,_1.
an rnterpo anon suc as a 
plate spline) a "homology map" • hypergeometric
means that one intends to refer tion ,
to its features as if they had the
something to do with valid bio ; cion with parameters, N, G and
logical explanations pertaining : n is the distribution of the
to the regions between the number of in
landmarks, about which we ; a simple random sample of
have no data. : size n (i.e., a random sample
• homomorphism without replacement in which
a function that preserve the op ; every subset of size n has the
erators associated with the : same chance of occurring)
specified structure. from a population of N ob
• homoscedasticity
"Same scatter." A scatterplot or
residual plot shows
homoscedasticity if the scatter
• jects of which G are "good."
: The chance of getting exactly
g good objects in such a
• sample is GCg x NGCng!
; NCn, provided g < = n, g < =
MIJ ... nu======== ___ 11
76
hyperplllne I hypothesis tutino II
G, and n g < = N G. (The
probability is zero otherwise.)
The expected value of the hy
pergeometric distribution is
n x GjN, and its standard er
ror is ((Nn)j(NI»V2 x (n X
GjN x (IGjN) ) V2 .
I • hypotenuse
I the side opposite the right angle
in a right triangle. The other
I two sides are called legs.
I • hypothesis testing
I statistical hypothesis testing is
formalised as making a decision
between rejecting or not reject
ing a null hypothesis, on the
basis of a set of observations.
Two types of errors can result
I from any decision rule (test):
rejecting the null hypothesis
when it is true (a Type I error),
I and failing to reject the null hy
pothesis when it is false (a Type
II error). For any hypothesis, it
I is possible to develop many dif
• hyperplane :
a kI dimensional subspace of I
a kdimensional space. A hy :
perplane is typically
characterised by the vector to
which it is orthogonal.
• hyperspace
a space of more than three
dimensions.
I ferent decision rules (tests).
: Typically, one specifies ahead of
time the chance of a Type I er
; ror one is willing to allow. That
chance is called the significance
level of the test or decision rule.
I For a given significance level,
one way of deciding which de
cision rule is best is to pick the
I one that has the smallest chance
of a Type IT error when a given
alternative hypothesis is true.
• . The chance of correctly reject
a generalisanon of the ldea of ; ing the null hypothesis when a
• hypersphere
a generalisation of the idea of a
sphere to a space of greater than
three dimensions.
volume to a space of more than given alternative hypothesis is
three dimensions.
II icoSahedron I implies, ~ i c a l imPlica; 77
true is called the power of the ~ tion ( (p AND q) OR ((NOT
test against that alternative. ; p) AND (NOT q» ).
• icosahedron
a polyhedron with 20 faces.
Usually refers to a regular
icosahedron, one of the Platonic
solids .
; • image
: the reflection of the preimage
I
: • imaginary axis
~ the yaxis of an Argand diagram.
. d ~ • imaginary number
.1 empotent .
. . I a complex number of the form
the element x III some algebraJ.c. . h . al d' t(
. . . . Xl were x IS re an l=sqr 
structure IS called Idempotent if ~ 1).
x*x=x. .
• identify
imaginary
to choose from a set or to name
Simplify:
cases in which the desired re I
suIt is present or true.
a)./=4 . J( I)(.) ' "
• identitiy transformation
a size change where k equals 1
• identity reflection
a reflection where the preimage
and the image are the same.
• iff
I
: • imaginary part
~ the imaginary part of a complex
I munber x+iy where x and y are
: real is y.
I
: • implies, logical implication
• iff, if and only if ~ logical implication is an opera
if p and q are two logical propo I tion on two logical proposi
sitions, then(p iff q) is a propo ~ tions. If p and q are two logical
sition that is true when both p : propositions,(p IMPLIES q) is
and q are true, and when both ~ a logical proposition that is true
p and q are false. If is logically ; if P is false, or if both P and q
equivalent to the proposition ( : are true. The proposition (p
(p IMPLIES q) AND (q IM ~ IMPLIES q) is logicallyequiva
PLIES p) ) and to the proposi ; lent to the proposition ((NOT
: p) ORq).
if and only if.
I
MAtbmuJncs======= II
subset I irulepnulmt, irukperulm&e II
• improper subset I • independent and identi
a subset that includes the entire I cally distributed
parent set.
• incenter
the point of concurrency of a
triangle's three angle bisectors.
• incircle
a collection of two or more ran
I dom variables {Xl, X2, . . . , }
I is independent and identically
distributed if the variables have
I the same probability distribu
I tion, and are independent .
I • independent variable
in regression, the independent
• included angle variable is the one that is sup
an angle formed between two I posed to explain the other; the
given sides of a triangle. term is a synonym for "explana
the circle inscribed in a given
figure.
• inclusive OR tory variable." Usually, one re
one or the other, or both; and/ I gresses the "dependent vari
able" on the "independent vari
able." There is not alwavs a
or
• incoming angle
the angle formed between the
path of an approaching object
(a billiard ball, a light ray) and
the surface it rebounds against
(a cushion, a mirror) .

I clear choice of the independent
variable. The independent vari
able is usually plotted on the
I horizontal axis. Independent in
I this context does not mea.., the
same thing as statistically inde
pendent .
• independent, indepen
dence
two events A and B are (statis
independent if the
I chance that they both happen
simultaneously is the product of
tl:ie chances that each occurs in
I dividually; i.e., if P(AB) =
P(A)P(B). This is essentially
equivalent to saying that learn
II =======MJJt __ ri&s
79
II iruJ,icaturrandmn variable I indirta;s"=rem&n==t=========
ing that one event occurs does every event determined by X
not give any information about I and Y
whether the other event oc : • indicator random variable
curred too: the condi tional the indicator [random variable]
probability of A given B is the of the event A, often written
same as the unconditional prob ; lA, is the random variable that
ability of A, i.e., P(AIB) = : equals unity if A occurs, and
P(A). Two random variables X zero if A does not occur. The
and Yare independent if all ; expected value of the indicator
events they determine are inde : of A is the probability of A,
pendent, for the P(A), and the standard error of
event {a < X < = b} IS mde I the indicator of A is (P(A) x (1
pendent of the event {c < Y < = : P(A)) V2. The sum lA + IB +
d} for all choices of a, b, c, and 1 C + ... of the indicators of a
d. A collection of more than two ; collection of events {A, B, C, ..
random variables is indepen . } counts how many of
dent if for every proper subset : events {A, B, C, ... } occur m
of the variables, every event a given trial. The product of the
determined by that subset of the ; indicators of a collection of
variables is independent of ev : events is the indicator of the
ery event determined by the intersection of the events (the
variables in the complement of ; product equals one if and only
the subset. For example, the : if all of indicators equal one).
three random variables X, Y, The maximum of the indicators
and Z are independent if every ; of a collection of events is the
event determined by X is inde : indicator of the union of the
pendent of every event deter events (the maximum equals
mined by Y event lone if any of the indicators
determined by X IS mdependent : equals one).
of every event determined by Y . .
and Z and every event deter ; • mdirect
mined by Y is independent of : a .that IS lffipOS
every event determined by X sible lffipraCtlcal t? be
and Z and every event I sured directly or phYSICallY, us 
mined by Z is independent of
MlJthenuJrics======= II
80
~ ~ ~ ~ = ~ ~ ~ = .
indirect proof I inst:ribetl1l1llJk "
ally calculated using a formula I • infinitesimal
or a known relationship. I a variable that approaches 0 as
• indirect proof
a limit.
a proof that begins by assum • inflection
ing the conclusion is not true ~ a point of inflection of a plane
and leads to a contradiction of ; curve is a point where the
either the assumption or a pre curve has a stationary tangent,
viously proved theorem. at which the tangent is chang
I ing from rotating in one di
• indirect technique
the method used to determine I rection to rotating in the
an indirect measurement.
oppostie direction.
• inductive reasoning
• initial side
I the side that the measurement
of an angle starts from ..
the process of observing data,
recognizing patterns, and maleI
ing conjectures about
• injection
generalisations. a onetoone mapping.
• inequality • inscribed
the statement that one quantity I (in a polygon or polyhedron)
Intersecting each side or face is less than (or greater than)
another.
• infinite
of a figure exactly once. Usu
I ally referring to circles in
scribed in polygons or spheres
inscribed in polyhedrons. The
I figure outside is circum
becoming large beyond bound. I
scribed around the inscribed
figure.
• inscribed angle
an angle formed by two chords
I of the circle with a common
endpoint (the vertex of the
angle).
II instance ofa sentence I intersecting p:es===========""8,,,1
• instance of a sentence ~ circle IS less than that of the
a situation where the statement ; radius.
is true
• integer
any whole number or its oppo
Site.
• integral coefficient
in the expression, 3x, 3 is the
coefficient. Integral coeffi
cients are coefficients that are
integers.
• intercepted arc
an arc of a circle whose end
points are marked by the sides
of a central angle or an lll
scribed angle.
Angle A = Angle B
• interior angles
; • interior of an angle
the convex set formed by an
I angle that measures less than
180 degrees.
I • internally tangent
(circles) Intersecting at exactly
one point, with one circle inside
I the other.
I • interpolate
determine a value within a set
of given values using a formula,
I rule, or reasonable estimation.
I • interpolation
given a set of bivariate data (x,
y), to impute a value ofy corre
~ sponding to some value of x at
; which there is no measurement
of y is called interpolation, if
I the value of x is within the range
of the measured values of x. If
the value of x is outside the
I range of measured values, im
I puting a corresponding value of
y is called extrapolation.
• interquartile range
angles between two lines cut by I
the interquartile range of a list
I of nwnbers is the upper quartile
minus the lower quartile.
a transversal.
• interior of a circle
the set of points whose dis I
tance from the center of the I • intersecting planes
planes that share a line
MR.thematics=================== II
I II
I fully be added or subtracted and
I that the mean is a representa
tive measure of central ten
I dency. Such data are common
I in the domain of physical sci
ences or engineering e.g.
lengths or weights .
• invar.iant
I an invariant, generally speaking,
the set of elements which are is a quantity that is unchanged
in both A and B. (even though its formula may
I have changed) when one
• intersection of two sets A
andB
A.B or A.,,B
The above operation is cOIIum.alive. associative, changes some inessential aspect
anddistribttive of a measurement. For in
AB=BA (AB)C=A.(BC) A(B+C)=AB+AC I stance, Euclidean distance is an
We note that if A. c B, thenAB =A.. Hence invariant under translation or
AA =.'1. {$} A = {$} AF= A rotation of one's coordinate sys
• intersection I tern, and ratio of distances in
the intersection of two or more I the same direction is an invari
sets is the set of elements that ant under affine transforma
all the sets have in common; the I tions. In the morphometrics of
elements contained in every one I triangles, the invariants of a
of the sets. The intersection of : particular transformation are
I
the events A and B is written the shape variables that do not
'1\ and B" and '1\B." C.f. union. I change under that transforma
• interval scale
tion.
• inverse
a form of conditional; if not p,
I then not q.
a characteristic of data such that
the difference between two val
ues measured on the scale has
the san1e substantive meaning!
significance irrespective of the
common level of the two val I
ues being compared. This im
plies that scores may meaning
I • inverse operation
examples of inverse operations
are addition and subtraction,
I multiplication and division, ex
II ===================MJJthemllti&s
II in17eNe sine, cosine, or tangent I isome; 83
tracting a root and raising to a I combinations of shapes;
power. shapes that may contain
• inverse sine, cosine, or
tangent
(of a number) The acute angle
whose sine, cosine, or tangent
is the given number.
• inverse variation
curved portions rather than
I straightline segments.
'.ISO
I acronym for the International
Standards Organisation, based
in Geneva, Switzerland. This
I body publishes specifications
for a number of standard pro
gramming languages. The
I specifications are arranged gen
a variation of the type xy = k;
the graph is a hyperbola with I
axes as asymptotes.
• irrational number
decimal number that never
ends, never repeats (Ex: pi)
• irrational number
a number that is not rational.
• irregular region
region whose boundary is not the
union of circular arcs or seg
ments
• irregular shapes
shapes that are not one of the
named geometric shapes or
~ .
••
I
erally to concur with those of
ANSI.
: • isogonal conjugate
~ isogonal lines of a triangle are
; cevians that are symmetric with
: respect to the angle bisector.
~ Two points are isogonal conju
; gates if the corresponding lines
to the vertices are isogonal.
• isometric drawing
I a type of drawing that shows
I three faces of a threedimen
sional object in one view. The
isometric drawing of a cube
I shows all the edges equal, but
each square face is represented
as a 60°120°60°120° rhom
I bus.
I • isometry
: an isometry is a transformation
of a geometric space that leaves
Mllthemntics==================== II
84
=======*
isosceles tetrahedron I isotropic II
distances between points un
changed. If the space is the Eu I
clidean space of a picture or an
organism, and the distances are
distances between landmarks,
ean translations, rotations, and • isosceles triangle
reflections. If the distances are I a triangle with at least two con
Procrustes between I gruent sides. If a triangle has
shapes, the isometnes the I exactly two congruent sides,
simplest case, landmarks two : they are called the legs and the
dimensions) are the rotatlons I anoIe between them is called the
Kendall's shape .. For. tn angle. The opposite
angles, these can be : the vertex angle IS called the
as ordinary rotatIOns of I base. The nonvertex angles are
Kendall's "spherical black called the base angles.
board."
• isosceles tetrahedron
a tetrahedron in which each pair
of opposite sides have the same
length.
• isosceles trapezoid
a trapezoid whose two nonpar
allel sides are congruent.
• isoscoles trapezoid
a trapezoid that has a pair of
I equiangular base angles
I • isotomic conjugate
two points on the side of a tri
angle are isotomic if
I equidistaru: from the.
of that side. 1\vo pomts mSlde
a triangle are isotomic conju
I d.
oates if the correspon lllg
b .
I cevians through these pomts
meet the opposite sides in
isotornic points .
• isotropic
I invariant with respect to di
rection. Isotropic errors have
the same statistical distribu
II =======MR.th_ti.cs
II jointprobability distribution 'I Kenda;'S""S""hR""ifJ""e
tion in all directions implying
equal variance and zero corre
lation between the original
variables (e.g., axis coordi
nates).
_ joint probability distri
bution
if Xl, X2, ... , Xk are ran
 Jordan matrix
; a matrix whose diagonal ele
: ments are all equal (and non
zero) and whose elements
; above the principal diagonal are
: equal to 1, but all other ele
I
. ments are O.
0 0 0 1 0 0 0
dom variables, their joint I
probability distribution gives
0
0
0
0 2
0 0 0 0
0 0 0 0
the probability of events de ..
termined by the collection of I
random variables: for any col
A:=
0
0 0
0 1
2 0 0 0
0 2 0 0
lection of sets of numbers
3 2 0 1 2
{AI, ... ,Ale}, the joint prob I
0 1 0 0 0 0
ability distribution deter I •
mines P( (Xl is in AI) and:  Joule
(X2 is in A2) and ... and (Xk a unit of energy or
is in Ak) ). _ jump discontinuity
_ joint probability function ; a discontinuity in a function
a function that gives the prob : where the left and righhand
ability that each of two or limits exist but are not equal to
more random variables takes ; each other.
at a particular value.
 Jordan curve
a simple closed curve.
;  justify
: give a logical explanation or in
formal proof of a mathemati
; cal situation, computation or
: property.
I
:  Kendall's shape space
the fundamental geometric con
I strUction, due to David Kendall,
: underlying geometric
morphometrics. Kendall's
MRthmuJri&s======= II
86
kilonIeml I II
shape space provides a com I sides are called the vertex
plete geometric setting for I angles. The angles between the
analyses of Procrustes distances pairs of noncongruent sides are
among arbitrary sets of land I called the nonvertex angles.
marks. Each point in this shape I • knight's tour
space represents the shape of a I a knight's tour of a chessboard
configuration of points in some : is a sequence of moves by a
Euclidean space, irrespective of I knight such that each square of
size, position, and orientation. the board is visited exactly once.
In shape space, scatters of :
points correspond to scatters of • knot
entire landmark configurations, . a ·curve in space formed by in
not merely scatters of single terlacing a piece of string and
landmarks. Most multivariate I then joining the ends together.
methods of geometric I
morphometrics are
linearisations of statistical I
analyses of distances and direc I
tions in this underlying space.
• kilometer
a unit of length equal to 1,000 I
meters .
• kinematics
a branch of mechanics dealing
with the motion of rigid bodies
without reference to their I
masses or the forces acting on
the bodies.
.1ONM "Ill" IMTZIW' .......
WI ..
,D ..... ' '_ AU· DlJllHI7fIt
STfJt ITDC AA4ItIIt .. ,,,.
• kite
I the crucial matrix for comput
ing the thinplate spline
interpolant between two land
I mark configurations. In this
I entry, k stand,s"for the number
a quadrilateral with exactly two I
distinct pairs of congruent con
secutive sides. The angles be
tween the pairs of congruent
of landmarks, for historical
II
reasons . The equation of the
thinplate spline has coeffi
cients Vlh, where h is. a vec
tor of the xor ycoordinates of
the landmarks in a target
fo;m, followed by three D's
(for two dimensional data,
four D's for threedimensional
data). The entries in the ma
trix L are wholly functions of
the starting or reference form
for the spline. Bending energy
is the upper kbyk square of
VI.
.la
lateral area
• latera recta
plural of lattice rectum.
• lateral area
87
* = = = = = = = = = = = = = = = = ~
~ • lateral faces
; the faces of the lateral surface
: of a prism, or a face of a pyra
~ mid that is not a base
~ • lateral surface
; the surface not included in the
: base(s)
I
: • latin square
~ an n X n array of numbers in
I which only n numbers appear.
: No number appears more than
I . I
: once In any row or co umn.
~ • latitude
I the angular distance of a point
~ on the Earth from the equator,
: measured along the meridian
I through that point.
the area of the lateral surface I
of a solid
• lateral edge
s
a segment whose endpoints are
corresponding points of a cylin I
dric solid's bases, or whose end I
points are the vertex of a conic
solid and a vertex of its base
~ • lattice points
; points in the coordinate plane
• lateral edge : with integer coordinates
(of a prism) The intersection of I
two lateral faces. : • latus rectum
I a chord of an ellipse passing
through a focus and perpen
M..themiJtUs======= II
88 I law ofla1lJenumhers II
dicular to the major axis of the I • law of contrapositive
ellipse.Plural: latera recta. I the type of valid reasoning that
• law of averages
I
I
I
I
I I
"=1.)
concludes the truth of a state
ment from the truth of its
I contrapositive .
• laW' of cosines
the theorem that, for any tri
o angle with angles of measure A,
B, and C, and sides of lengths
; a, and c (a opposite A, b op
poSIte B, and c opposite C), c2
I = a2 + b2  lab cos C.
the Law of Averages says that
the average of independent ob I • law of large numbers
servations of random variables the Law of Large Numbers says
that have the same probability that in repeated, independent
distribution is increasinoIy I trials with the same probability
likely to be close to the P of success in each trial, the
value of the random variables percentage of successes is in
as the number of observations I creasingly likely to be close to
grows. More precisely, if X!, I the chance of success as the num
X2 X3 are
0 d d ber of trials lOncreases. More
, ,..., III epen ent
random variables with the same precisely, the chance that the
probability distribution, and I percentage of successes differs
E(X) is their common expected from the probability p by more
value, then for every number E than a fixed positive amount, E
> 0, P{ I (Xl + X2 + ... + Xn)j I > 0, converges to zero as the
n E(X) I < E} converges to number of trials n goes to in
100% as n grows. This is fInity, for every number e > O.
equivalent to saying that the I Note that in contrast to the dif
sequence of sample means Xl, I ference between the percentage
(XI+X2)j2, (Xl+X2+X3)j3,. of successes and the probabil
.. converges in probability to I ity of success, the difference
E(X). between the number of suc
cesses and the expected number
II =======MRthmMtiu
II law of sines I likelihood ratio test
of successes, n X p, tends to
grow as n grows. The follow
ing tool illustrates the law of
large numbers; the button
toggles between displaying the
difference between the number
of successes and the expected
number of successes, and the
difference between the percent
age of successes and the ex
pected percentage of successes.
The tool on this page illustrates
the law of large numbers.
• law of sines
the theorem that, for any tri
angles with angles of measure
A ~ B ~ and C ~ and sides of lengths
a ~ b ~ and c (a opposite A ~ b op
posite B ~ and c opposite C), sin
sin sin A = sin B = sin C
abc
• least common multiple
the least common multiple of a
set of integers is the smallest
integer that is an exact multiple
of every number in the set.
• least upper bound
the least upper bound of a set
of numbers is the smallest num
ber that is larger than every
member of the set.
89
*=================
~ • leastsquares estimates
; parameter estimates that
: minimise the sum of squared
~ differences between observed
; and predicted sample values.
; • leg of a right triangle
: a side of a right triangle that
~ include the 90 degree angle
I
: • legs
I (of an isosceles triangle), the
: two congruent sides of a non
~ equilateral isosceles triangle.
I
.• lemata
~ plural of lemma.
I. lemma
~ a theorem whose importance is
: primarily as part of a proof of
I another theorem.
I • likelihood ratio test
; a test based on the ratio of the
: likelihood (the probability or
~ density of the data given the
; parameters) under a general
: model, to the likelihood when
~ another, specified hypothesis is
; true. Many of the commonly
: used statistical tests are likeli
~ hood ratio tests, e.g., the ttest
I for comparisons of means,
: Hotelling's T2, and the analysis
If ·
: 0 vanance Ftest.
MAthemtltics======= II
90
limit II
=========*
• limit
L I\.
the actual area of a region
• line
E E )
r v
• line of symmetry
an undefined term in most de
ductive systems. A straight ar I
rangement of infinitely many I
points. A line has no thickness
and is therefore considered one
dimensional.
the line of reflection of a figure
I having reflectional symmetry. y
i
t
I • line perpendicular to a
plane
v
a line perpendicular to every line
I in the plane that it intersects (or
anyone of them)
• line of best fit I
given a collection of points, a :
line that passes closest to all of
them, as measured by some .
given criterion.
A
/
I E
/
/
/1
.'
i I
L __________ . ___ j
.I
I • line segment
• line of reflection two points (the endpoints of the
the line over which every point I line segment) and all the points
of a figure is moved by a reflec I between them that are on the
tion. line containing them. The line
segment connects the points.
I The measure of a line segment
I is its length.
II =======MJJdlmulriu
lilinesyrnmetry Ilineartenn
of
anequa;===========9=1
• line symmetry
a figure has line symmetry if
there is a least one line that di
vides the figure into two parts
that are mirror images of each
other.
• linear operation
; suppose f is a function or op
: eration that acts on things we
shall denote generically by the
; lowercase Roman letters x and
: y, Suppose it makes sense to
multiply x and y by numbers
• linear association
two variables are linearly asso I (which we denote by a), and
that it makes sense to add
ciated if a change in one is as I
sociated with a proportional things like x and y together, We
change in the other, with the I say that f is linear if for every
. f' al I number a and every value of x
same constant 0 propornon  d fi hi h C() d C )
· h h h f : an y or w c l' x an 1(y are
lty t roug out t e range 0 I d fi d (') C( ). d fi d
Th l
· e me , 1 l' a x x IS e me
measurement, e corre anon: d al £:() d ") £:(
coefficient measures the degree I an equ s aX x, an (u x
of linear association on a scale : + y) is defined and equals f(x)
of 1 to 1. + f(y) .
• linear combination
a sum of values each multiplied
by some coefficient . A linear
combination can be expressed
as the inner product of two vec
tors, one representing the data
and the other a vector of coelli
cients.
• linear equation
ax + By + C = 0
• linear function
a function that, when applied to
consecutive whole numbers,
generates a sequence with a
constant difference between
consecutive terms .
• linear pair
; 2 supplementary adjacent
: angles whose noncommon
sides form a line.
• linear pair of angles
; two adjacent angles whose dis
tinct sides lie on the same line.
<
Angles 1 and 2 form a
Linear Pair
)
I • linear term of an equation
; the term a but no
: exponent m an equauon; ex
I ample: By in a linear equation
.MiJt"'nu======= II
",,92===========* lineartransfOrmatUm IlocatUm II
• linear transformation
in multivariate statistics, a lin
ear transformation is the con
struction of a new set of vari
abIes that are all linear combi
nations of the original set. In
geometric morphometrics, one
linear transformation takes
Procrustesfit coordinates to
I ments, usually bits of geometry
I or whole functions, that can be
added together and can be mul
I tiplied by real numbers in an
I intuitive way. The points of a
plane don't form a linear vec
tor space (what is "five times a
I point"?), but lines segments
connecting all the points to the
origin do form such a space. partial warp scores; another
takes them to relativewarp I
scores. A linear transformation
• lines of sight
I lines from an eye to what it sees
that show perspective 'and what
size to draw it
of a matrix A can be written in
the form y = Ax, where y is the ~
resulting linear combination of ;
x, a column vector, with the
• loading
rows of A.
• linear vector space
x
j
I the correlation or covariance of
a measured variable with a lin
ear combination of variables. A
I loading is not the same as a co
I efficient. In general, coefficients
supply formulas for the compu
I tation of scores whereas load
I ings are used for the biological
interpretation of the linear com
bination .
• local behaviour
I a description of the values of a
function or relation within a
small interval of the indepen
I dent variable.
in morphometrics, the most
common kdimensional linear
vector space is the set of all real I
kdimensional vectors, includ
ing all sums of these vectors and
their scalar multiples. More
generally, but informally, a lin
ear vector space is a set of ~ l e 
I • location
one of the four main descrip
tion of a point
II =======MlJthenuuies
111OCRtion,
• location, measure of
I sion. An argument is valid if the
I conclusion has been arrived at
through deductive reasoning.
a measure of location is a way
of summarising what a "typi
cal" element of a list isit is a I
onenumber summary of a dis
• logistic regression
I this relates to an experimental
I design for predicting a binary
tribution.
• locus
the path of a moving point; the
set of all points in a plane satis
fying some given condition or
property.
• logarithm
categorical (yes/no) olltome on
the basis of predictor variables
I measured on interval scales. For
each of a set of values of the
a logarithm of a number is the I
exponent to which a given base
must be raised to produce the
given number.
predictor variables, the out
I comes are regarded as repre
senting a binomial process,
with the binomial parameter 'p'
I depending upon the value of the
predictor variable. The model
ling accounts for the logarithm
I of the odds ratio as a linear
I function of the predictor vari
able. Fitting is via a weighted
• logarithmic growth
a set of values which are ap
proximated by an equation of
the form y = log b x.
• logarithmic notation
use of the symbols "log" or "in"
in context.
I 1 .
eastsquares regreSSIOn
I method. randomisation tests
for this purpose have been de
veloped by Mehta & Patel.
• longitudinal study
I a study in which individuals are
followed over time, and corn
reasonmg.
pared with themselves at differ
• logic
the study of the formal laws of
• logical argument I ent times, to determine, for ex
a set of premises followed by ample, the effect of aging on
statements, each of which relies I some measured variable. Lon
on the premises or on previous I gitudinal studies provide much
statements, and ending with a more persuasive evidence about
final statement called a conclu
Mathemlltics==================== II
94
=========*
loJl1er bountl I mRBie tour II
the effect of aging than do cross I numbers. Lo = 2, L} = 1, Ln = Ln_
sectional studies. I } + Ln_r
• lower bound I • lune
any number below which a the portion of a sphere between
function value may approach I two great semicircles having
but not pass. I common endpoints (including
• lowest common denorni
nator
the smallest number that is ex
actly divisib.1e by each denomi
nator of a set of fractions.
.loxodrome
the semicircles).
• magic square
a square array of n numbers
I such that sum of the n numbers
in any row, column, or main di
agonal is a constant (known as
I the magic sum).
on a sphere, a curve that cuts
all parallels under the same I
angle.
4
3
8
9
5
1
2
7
6
I • magic tour
if a chess piece visits each square
I of a chessboard in succession,
• ltetromino I this is called a tour of the chess
a tetromino in the shape of the board. If the successive squares
letter L. I of a tour on an n X n chess
I board are numbered from 1 to
• lucas number
n A 2, in order, the tour is called
a member of the sequence 2, 1, I a magic tour if the resulting
3, 4, 7, ... where each number is I square is a magic square.
the sum of the previous two
II =======MAtlmnRtics
II magnitude I MannWhitney test
• magnitude
the value of a number; its dis
tance from the origin
• magnitude of a rotation
the amount of rotation in de
grees
• magnitude of a translation
the distance between any point
and its image
• major arc
an arc whose endpoints form
an angle over 180 degrees with
the center of the circle; written
the extra letter is used to dis
tinguish it from a minor arc.
• major axis
the major axis of an ellipse is
it's longest chord.
Reference srs1ltml
• malfatti circles
three equal circles that are mu
tually tangent and each tangent
to two sides of a given triangle.
95
*================
~ • manipulatives
; objects that can be arranged,
: built, and moved around by
~ hand,
~ • MannWhitney test
; this is a test of difference in
: location for an experimen
~ tal design involving two
I samples with data mea
: sured on an ordinal scale or
~ better. The test statistic is
~ a measure of .<?:rdinal prece
; dence. For each possible
: pairing of an observation in
~ one group with an observa
; tion in the alternate group,
: the pair is classified in one
~ of three ways according to
; whether the difference is
: positive, zero or negative;
~ the numbers in these three
; categories are tallied over
: the randomisation set. The
~ randomisation set is the
I same as that for the Pitman
~ permutation test. This test
: is generally recommended
~ for comparisons involving
; ordinalscale data but is not
: confined to this scaletype.
~ An equivalent formulation
; of the test, based upon
: ranking the data and sum
~ ming ranks within groups,
; is the Wilcoxon test.
MsthmuJties======= II
OVA I mmeimumliltelihootlestimtlte II
.MANOVA
see multivariate analysis of vari
ance.
.mapping
making a transformation
• margin of error
world situation is abstracted to
I a model, the related math
ematical problem is posed and
I solved, and the mathematical
I solution is interpreted back into
the realworld situation as a
solution to the realworld prob
lem.
I • mathematical notation
correct use of labels, symbols,
and abbreviations in a math
a measure of the uncertainty in I
an estimate of a parameter;
unfortunately, not everyone
agrees what it should mean.
The margin of error of an esti
I ematics contett.
mate is typically one or two
times the estimated standard I • matrix
error of the estimate.
• Markov's inequality
arrangement of pixels
• maximum
the largest of a set of values.
• maximum likelihood
estimate
the maximum likelihood esti
mate of a parameter from data
is the possible value of the pa
rameter for which the chance of
observing the data largest. That
is, suppose that the parameter
for lists: If a list contains no
negative numbers, the fraction I
of numbers in the list at least
as large as any given constant
a>O is no larger than the arith I
metic mean of the list, divided I
by a. For random variables: if a
random variable X must be I
nonnegative, the chance that X I
exceeds any given constant a>O
is no larger. than the expected
value of X, divided by a.
I is p, and that we observe data
x. Then the maximum likeli
hood estimate of p is estimate
• mathematical model
I p by the value q that makes
a mathematical q,bject (such as
a geometric graph,
table, or equation) representing I
a realworld situation. In math
ematical modeling the real
P(observing x when the value
of p is q) as large as possible.
For example, suppose we are
trying to estimate the chance
that a (possibly biased) coin
II =======MA.themsries
II mean I meamremmttype
97
* = = ~ ~ = = = = ~ ~ = =
lands heads when it is tossed. ~ symbols, if X is an estimator of
Our data will be the number of ; the parameter t, then
times x the coin lands heads in
n independent tosses of the coin.
The distribution of the number
of times the coin lands heads is
binomial with parameters n
(known) and p (unknown). The
chance of observing x heads in
n trials if the chance of heads in
a given trial is q is ne
x
qx( lq)n
x. The maximum likelihood es
timate of p would be the value
of q that makes that chance
largest. We can find that value
of q explicitly using calculus; it
turns out to be q = xjn, the frac
tion of times the coin is ob
served to land heads in the n
tosses. Thus the maximum
likelihood estimate of the
chance of heads from the num
ber of heads in n independent
tosses of the coin is the observed
fraction of tosses in which the
coin lands heads.
; • mean, arithmetic mean
: the sum of a list of numbers,
~ divided by the number of num
i bers. See also average.
I.meaning
: a version of a conditional that
I
: defmes a term, where the term
I is in the antecedent.
I. measure
~ the amount of openness in an
: angle
I
: • measure of an angle
I the smallest amount of rotation
~ necessary to rotate from one
: ray of the angle to the other,
~ usually measured in degrees.
I • measure of an arc
~ the measure of minor arc or
: major arc is the measure of its
~ central angle.
~ • measurement type
; this is a distinction regarding
• mean : the relationship between a phe
average ~ nomenon being measured and
• mean squared error i the data as recorded. The main
the mean squared error of an : distinctions are concerned with
estimator of a parameter is the ~ the meaningfulness of numeri
expected value of the square of ~ cal comparisons of data (nomi
the difference between the es . nal scale versus ordinal scale
I . al
timator and the parameter. In : versus mterv scale versus ra
Mathematics================= II
98 ;w mangle r method ofcomptlrison II
tio scale this is known as I in the list after sorting the list
Stevens' typology), whether the I into increasing order. IT the list
scale of the measurements has an even number of entries,
(other than nominal scale mea I the median is the smaller of the
surements) should be regarded I two middle numbers after sort
as essentially continuous or dis ing. The median can be esti
crete, and whether the scale is mated from a histogram by
bounded or unbounded. I fmding the smallest number
• medial triangle
the triangle whose vertices are
the midpoints of the sides of a
given triangle.
A
8
c
such that the area under the his
togram to the left of that num
I ber is 50%.
2. (of a triangle), a line segment
connecting a vertex to the mid
I point of the opposite side.
I • member of a set
something is a member (or el
ement) of a set if it is one of
I the things in the set.
I • mersenne number
I a number of the form 2pl
where p is a prime.
• size • mersenne prime
a SIZe measure based on the re I a Mersenne number that is
peated . median of prime.
interlandmark distances. Used·
in resistantfit methods. ; • method of comparison
: the most basic and important
• median method of determining
1. "Middle value" of a list. The ; whether a treatment has an ef
smallest number that .at : fect: compare what happens to
least half the numbe.rs m the individuals who are treated (the
are no greater than It. IT the I treatment group) with what
has an odd number of entrIes, : happens to individuals who are
the median is the middle entry not treated (the control group).
I
1\ metric space I minor lire
99
• metric space • minimal path
a space and a distance function ; the path of shortest length, as
defined on every pair of points : when fmding the shortest path
that meets the requirements of from one point to another by
the definition of "metric" ; way of a fIxed line.
above.
; • minimalchange sequence
• midp : exploration of a randomisation
proposed by H.O Lancaster, distribution is such a sequence
and further promoted by G.A. I that the successive
Barnard. This is a tail defIni : randomisations differ is a
tion policy that the alpha value simple way. In the context of a
should be calculated as the I ranodmisation test this can
sum o( the proportion of the mean that the value of the test
tail for data strictly more ex : statistic for a particular
treme than the outcome, plus randomisation may be calcu
one half of the proportion of ; lated by a simple adjustment to
the distribution corresponding : the value for the preceding
to the exact outcome value. randomisation ..
This gives an unbiased esti
mate of alpha.
• midpoint
(of a line segment) The point
on the line segment that is the
same distance from both end
points.
• minimax strategy
; in game theory, a minimax
: strategy is one that minimises
one's maximum loss, whatever
; the opponent might do (what
: ever strategy the opponent
might choose).
Ods I ° °
• mt egment .• mtmmum
1.( of a trapewid), the line seg I the smallest of a set of values.
ment connecting the midpoints I • minor arc.
of the two nonparallel sides.· h d·
2 (f . gl) lin . an arc w ose en pomts form
. 0 a t:
nan
a . de an angle less than 180 degrees
t e ml pomts 0 ; with the center of the circle.
two Sl es :
MathBm4tics======= II
100
• minor axis
the minor axis of an ellipse is
its smallest chord.
.mira
minor IlI4ds I monic polynomi,lll II
~ a representation of a math
i ematical relationship or situ
: ation.
I
: • modulo
I the integers a and b are said to
I be congruent modulo m if ab
is divisible by m .
• modus ponens
the type of valid reasoning that
uses "if P then Q' and the state
ment P to conclude that Q must
be true.
• modus tollens
the type of valid reasoning that
uses "if P then Q" and state
ment "not Q' to conclude that
a plastic device which is used I
to determine and complete
symmetries by reflecting im
ages and allowing the user to
also see through the reflect
ing surface.
I "not P" must be trUe.
I • moment
the kth moment of a list is the
average value of the elements
I raised to the kth power; that is,
if the list consists of the N ele
• mixed variation
variation that contains both di
rect and inverse variation.
• mode
ments xl, x2, ... ,xN, the kth
for lists, the mode is a most
common (frequent) value. A I
list can have more than one I
mode. For histograms, a mode
is a relative maXImum
I moment of the list is ( xlk' +
x2k + xNk )fN. The kth mo
ment of a random variable X is
the expected value of Xk,
I B(Xk).
("bump").
• model
to create, using concrete ma
terials, drawings or symbols;
I ., monic polynomial
a polynomial in which the coef
ficient of the term of highest
I degree is 1.
II =======MJJthem4tics
• monochromatic triangle . randomisation set, sampled
a triangle whose vertices are all ; without replacement, and using
colored the same. : the values of the test statistic to
generate an estimate of the
• monohedral tiling ; form of the full randomisation
a tessellation in which all tiles
are c;ongruent.
• monomial
an algebraic expression consist
ing of just one term.
• monotone
a sequence is monotone if its
terms are increasing or decreas
: distribution. This procedure is
in contrast to the bootstrap pro
I cedure in that the sampling of
the randomisation set is with
: out replacement. An advantage
I of the MonteCarlo test over the
bootstrap is that with succes
: sive resamplings it converges to
the gold standard form of the
. . ; exact test. An effective necessity
• . : for the MonteCarloprocedure
mg.
a funct10n IS monotone if It only I is a source of random codes or
or only f an effective pseudorandom
Increases monotOnIcally (IS : generator.
monotonic increasing) if x > y, .
implies thatf(x) > = f(y). A ; • morphometrlcs
function f decreases monotoni . from the Greek: "morph,"
cally (is monotonic decreasing) I meaning "shape," and
: "metron " meaning "measure
if x > y, implies thatf(x) < = I '
f(y),. A function f is strictly : ment." Schools of
monotonically increasing if x > I morphometrics are
y, implies thatf(x) > f(y), and characterised by what aspects of
strictly monotonically decreas : biological "form" they are con
ing if if x > y, implies thatf(x) cerned with, what they choose
< f(y). ; to measure, and what kinds of
: biostatistical questions they ask
• MonteCarlo test . of the measurements once they
named after the famous SIte of . are made. The methods of this
gambling. casinos. A glossary emphasise configura
carlo test mvolves generaong a tions of landmarks from whole
random subset of the;
Msth_ties======= II
102 mse(x) =;( (#&')2) I multinomial tlistributiun II
organs or organisms analysed ~ of the bias and SE of the esti
by appropriately invariant bio ; mator: MSE(X) = (bias (X) )2
metric methods (covariances of : + (SE(X))2.
taxon, size, cause or effect with I • multimodal distribution
position in Kendall's shape I a distribution with more than
space) in order to answer bio lone mode. The histogram of a
logical questions. Another sort multimodal distribution has
of morphometries studies tis I more than one "bump."
sue sections, measures the den
sities of points and curves, and
.4
uses these patterns to answer I ••
••
questions about the random
processes that may be control
ling the placement of cellular I 4
structures. A third, the method
of "allometry," measures sizes
of separate organs and asks
questions about their correla
I • multinomial
I an algebraic expression consist
ing of 2 or more terms.
tions with each other and with
measures of total size. There
• multinomial distribution
are many others.
consider a sequence of n in
• mse(x) = e( (Xt)2 ) I dependent trials, each of
the MSE measures how far which can result in an out
the estimator is off from what come in any of k categories.
it is trying to estimate, on the I Let pj be the probability that
average in repeated experi each trial results in an out
ments. It is a summary mea come in category j, j = 1,2, ..
sure of the accuracy of the es I ., k, so pI + p2 + ... + pk =
timator. It combines any ten 100%. The number of out
deney of the estimator to comes of each type has a mul
overshoot or undershoot the I tinomial distribution. In par
truth (bias), and the variabil I ticular, the probability that the
ity of the estimator (SE). The n trials result in nl outcomes
MSE can be written in terms I of type 1, n2 outcomes of type
II =======MathimuJries
II multiple I 103
2, . . . , and nk outcomes of • multiplicity in hypothesis
type k is n!/(nl! x n2! x . . .; tests
X nk!) x plnl X p2n2 X .•. : in hypothesis testing, if more
X pknk, ifnI, ... ,nk are non than one hypothesis is tested,
negative integers that sum to ; the actual significance level of
n; chance is zero other : the combined tests is not equal
Wlse. to the nominal significance
• multiple
I level of the individual tests .
the integer b is a multiple of the  multivariate analysis of
integer a if there is an integer d variance
such that b=da. MANOVA. An analysis ofvari
_ multiple discriminant ance of two or more dependent
analysis ; variables considered simulta
discriminant analysis involving neously.
three or more a prioridefmed :  multivariate data
groups. I a set of measurements of two
_ multiple regression ; or more variables per indi
the prediction of a dependent vidual.
variable by a linear combination :  multivariate
of two or more independent morphometries
variables using leastsquares ; °a term historically used for the
methods for parameter estima : application of standard multi
tion. variate techniques to measure
; ment data for the purposes of
 multiplication rule
the chance that events A and : morphometric analysis. Some
B both occur (i.e., that event
AB occurs), is the conditional
probability that A occurs given
that B occurs, times the un
conditional probability that B
occurs.
what confusing now as any
I morphometric technique must
: be multivariate in nature.
I
:  multivariate multiple
I •
: regression
I the prediction of two or more
dependent variables using two
MIIthernatics======= II
!!::!lO!!::!4=====!!::!m!!::!ul!!::!tm!!::!' !!::!'I1II'!!::!Ut'!!::!te!!::!regremon: *". I negstiPe binomiRldistrihution II
or more independent vari ~ • nearly normal distribution
abies. ; a population of numbers (a list
• . : of numbers) is said to have a
• multivanate regression I
. nearly normal distribution if the
the prediction of two or more .
I histogram of its values in stan
dependent variables using one
independent variable.. : dard units nearly follows a nor
. ~ mal curve. More precisely, sup
• nadir I pose that the mean of the list is
the point on the celestial : JL and the standard deviation of
spehere in the direction down ~ the list is SD. Then the list is
wards of the plumbline. I nearly normally distributed if,
• nagel point for every two numbers a < b,
in a triangle, the lines from the the fraction of numbers in the
vertices to the points of contact I list that are between a and b is
of the opposite sides with the I approximately equal to the area
excircles to those sides meet in under the normal curve be
a point called the Nagel point. I tween (a JL)/SD and (a JL)/SD.
/'
• natural number
anyone of the numbers 1,2,3,
4,5, ....
• navigational system
I • negation
I (of a statement) A statement
: that is false if the original state
~ ment is true, and true if the
; original statement is false. The
: negation can usually be made by
~ appropriately adding the word
I not to the statement, or by pre
: ceding the statement with the
phrase "It is not the case that. .
I .. " A double negation of a state
ment is a negation of the nega
tion of that statement.
compass directions or bearings I
in a variety of formats.
• negative binomial distribu
tion
consider a sequence of inde
pendent trials with the same
II negllti:"enumbQ' I ninepointcircle .. 105
probability p of success in ~ together, which gives the ex
each trial. The number of tri ; pression for P(N =k) above.
als up to and including the rth
success has the negative Bino
mial distribution with param
eters n and r. If the random
variabJe N has the negative
binomial distribution with
parameters nand r, then
P(N=k) = k1Cr1 x pr x (1
p)kr, for k = r, r+ I, r+2, ...
, and zero for k < r, because
there must be at least r trials
to have r successes. The nega
; • negative number
: a number smaller than O.
I
: • net
~ a twodimensional pattern that
; you can cut out and fold to form
: a threedimensional figure.
I
: • network
~ a group of nodes and arcs
~ • ngon
; a polygon with n sides
tive binomial distribution is I • nine point center
derived as follows: for the rth : in a triangle, the circumcenter
success to occur on the kth ~ of the medial triangle is called
trial, there must have been r ~ the nine point center.
1 successes in the first k 1 tri . . . .
als and the kth trial must re ~ • rune pomt circle
suIt in success The chance of ; in a triangle, the circle that
the former is cite chance of r : passes through the midpoints of
1 successes in k 1 independent ~ ~ e sides is called the nine point
trials with the same probabil I CIrcle.
ity of success in each trial,
which, according to the Bino
mial distribution with param I
eters n=k1 and p, has prob
ability k1Cr1 x prl x (1
p)kr. The chance of the lat I
ter event is p, by assumption.
Because the trials are inde
pendent, we can find the I
chance that both events occur I
by multiplying their chances
M/lthemtI,tics=======:;;;;;;;;;:' II
II
• no causation without
manipulation
alpha criterion level, other
; wise not ('nonsignificant').
: The commonest conventional
I values for the nominal alpha
I criterion level are 0.05 and
a slogan attributed to Paul
Holland. If the conditions
were not deliberately manipu
lated (for example, if the situ
ation is an observational study
rather than an experiment), it I
is unwise to conclude that :
there is any causal relation I
ship between the outcome and I
the conditions. See post hoc
ergo propter hoc.
• node
0.01.
• nominal scale
this is a type of measurement
scale with a limited number of
possible outcomes which cannot
be placed in any order repre
I senring the intrinsic properties
a description of a point in a net I
work where it is possible for
two different segments to share
the same endpoints
of the measurements. Ex
amples : Female versus Male;
the collection of languages in
which an international treaty is
published.
• nomograph
• nominal alpha criterion
level
I a graphical device used for com
I putation which uses a straight
edge and several scales of num
bers.
a publicly agreed value for
typel error, such that the I
outcome of a statistical test is I
classified in terms of whether • nonagon
I a ninesided polygon.
the obtained value of alpha is :
extreme as compared with I
this criterion level. The fine :
detail of the comparison in
vol ves the tail definition I
policy. The outcome is classi
fied as showing statistical sig
nificance ('significant') if the I
outcome has low alpha as
compared with the nominal
0cId_
II nf11'Ull9onal number 11I01IIPfJ//rametric
• nonagonal number
a number of the form n(7n5)/
2.
.nonary
associated with 9
• nonconstant rate of
change
set of data or table of values
in which the amount of the
dependent variable does not
change by a constant value as
the value of the independent
variable changes by a constant
value.
• nonlinear association
; the relationship between two
: variables is nonlinear if a
change in one is associated with
; a change in the other that is
: depends on the value of the
first; that is, if the change in the
; second is not simply propor
: tional to the change in the: first,
independent of the value of the
I first variable.
I • nonoverlapping regions
regions that don't share interior
: points
I •
: • nonparametric test
• hi h all I a number of statistical tests
a set 0 POInts In c 0 were devised, mostly over the
segments connectIng POInts of. . d 19301960 . h th
th li
. I . th . perlO , Wit e
e set e entlre y In e set; I °fi b' . f b
synonym: concave.
• noneuclidean geometries
hyperbolic geometry
a geometry in which, through
a point not on a line, there are
infinitely many lines parallel
to the given line.
• noneuclidean geometry
solid geometry
• nonincluded side
the side of a triangle that is not
included by 2 given angles
. spec! !C 0 Jectlve 0 ypass
; ing assumptions about sam
: pIing from populations with
data supposedly conforming
; to theoretically modeled sta
: tistical distributions such as
the normal distribution. Sev
; eral of these tests were
: explictly concerned with ordi
nalscale data for which mod
I eling based upon continuous
functions is clearly inappro
: priate. These tests are implic
itly rerandomisation tests.
II
108 nonpersp:edrtllwing I1UJrmtJl4ppyoximation II
• nonperspective drawing I ciable, the survey suffer from
a threedimensional drawing I large nonresponse bias.
that doesn't use perspective
I • nonresponse
• nonregular shape in surveys, it is rare that every
a shape that does not have all lone who is "invited" to partici
sides congruent and all angles I pate (everyone whose phone
. congruent. number is called, everyone who
• nonresponse bias
is mailed a questionnaire, ev
I eryone an interviewer tries to
stop on the street . . . ) in fact
responds. The difference be
I tween the "invited" sample
sought, and that obtained, is the
in a survey, those who re
spond may differ from those I
who do not, in ways that are
related to the effect one is try
ing to measure. For example, I
a telephone survey of how
many hours people work is
likely to miss people who are I
working late, and are there
fore not at home to answer the
phone. When that happens, I
nonresponse.
• nonrigid transformation
a transformation that does not
preserve the size and shape of
the original figure.
• nonvertex angles
the survey may suffer from I (of a kite) The two angles be
nonresponse bias.
I tween consecutive noncongru
Nonresponse bias makes the ent sides of a kite.
result of a survey differ sys I
tematically from the truth. • normal
I perpendicular
• nonresponse rate
the fraction of nonresponders ~ • normal approximation
in a survey: the number of I the normal approximation to
nonresponders divided by the : data is to approximate areas
number of people invited to ~ under the histogram of data,
participate (the number sent I transformed into standard
questionnaires, the number of ~ units, by the corresponding
interview attempts, etc.) If the : areas under the normal curve.
nonresponse rate is appre ~ Many probability distribu
tions can be approximated by
II =======MRthemtlBes
II nomudcurve I nomuddistributUm
a normal distribution, in the
sense that the area under the
probability histogram is close
to the area under a corre
sponding part of the normal
curve. To fmd the correspond
ing part of the normal curve,
the range must be converted
to standard units, by subtract
ing the expected value and di
viding by the standard error.
For example, the area under
the binomial probability his
togram for n = 50 and p =
30% between 9.5 and 17.5 is
The tool on this page illus
; trates the normal approxima
: tion to the binomial probabil
ity histogram. Note that the
; approximation gets worse
: when p gets close to 0 or 1,
and that the approximation
improves as n increases.
I • normal curve
: the normal curve is the famil
iar "bell curve:," illustrated on
I this page. The mathematical
expression for the normal
: curve is y = (2xpi)V2Ex2/2,
where pi is the ratio of the cir
To use the normal ap ; cumference of a circle to its
proximation, we transform d· (3 14159265 )
d
: lameter. . .. ,
the endpoints to standar and E is the base of the natu
units, by subtracting the ex ; rallogarithm (2.71828 ... ).
pected value (for the Bino Th al .
. I d . bl : e norm curve IS symmet
mla ran om vana e, n X p = I .. d h .  0 d
15 fi th I f d
. nc aroun t e pomt x  ,an
or ese va ues 0 nan· .. c. al f
)
d di . di th ul b I pOSItiVe lor every v ue 0 x.
p an Vi ng e res t .y : The area under the normal
the error (for a BI curve is unity, and the SD of
nOffilal, (n x p x (lp)) 1/2 = ; the normal curve suitably de
3.24 for these values of n : fined, is also Many (but
p). The area normal approXI not most) histograms con
mation is the area under the ; d . d d' .
I b (9 5
verte Into stan ar units,
norma curve etween .: . I c. 11 th
15)/3.24 = 1.697 and (17.5 I approXimate y 10 ow e nor
5)/3 24 0 2 h
.; mal curve.
1 . =.77; t at area IS I
73.5%, slightly smaller than : • normal distribution
the corresponding area under I the normal distribution is a
the binomial histogram. See ; theoretical distribution appli
also the continuity correction. : cable for continuous interval
Msth_tUs========== II
lIO
scale data. It is related math • nth term .
ematically to the binomial and ; the number that a function rule
chisquare distributions and to : generates as output for a count
several named sampling dis ing number n.
tributions (including I • nuisance parameters
Student's t, Fisher's. F, parameters of a model that
Pearson's r); these samplmg : must be fit but that are not of
distributions are the interest to the investigator. In
of paramet.nc ; morphometrics, the
statlsical mfernece to WhICh : for translation and rotation are
rerandomisation statistics usually nuisance parameters.
are an alternative. ; th .
. • null hypo eslS
• normalise ; in order to test whether a sup
to normalise a ob : posed interesting pattern exists
ject is to transform It so in a set of data, it is usual to
some function of its coordI I propose a null hypothesis < that
nates or other parameters has : the pattern does not exist. It is
a prespecified value. For ex the unexpectedness of the de
ample, vectors are gree of departure of the ob
normalised by transformation . served data, relative to the pat
into unit vectors, which have tern expected under the null
length one. hypothesis, which is examined
• not, negation, logical ; by the measure alpha. Refer
negation I ence to a null hypothesis is com
the negation of a logical propo mon between rerandomisation
sition p, NOT p, is a and parametric statis
tion that is the logical opposite ; tiCS.
of p. That is, if P is true, NOT : • null model
p is false, and if.p is false, NOT .the simplest model under con
p is true. Negation takes prece I sideration. The null model for
dence over other logical opera : shape is the distribution in
tions. Kendall's shape space that
I arises from landmarks that are
II =========MlJthtmUlM
111
II null set I oblique triangle
*================
I
:  oblique
distributed by independent cir I
cular normal noise of the same I
variance in the original
digitising plane or space and I
drawn from a single, homoge I
neous population. It is exactly
analogous to the usual assump
tion of "independent identically I
distributed error terms" in con
ventionallinear models (regres
sion, ANOVA) .
~ at an angle that is not a mul
; tiple of 90 degrees.
 null set
a set with nothing in it
 number line
a line on which each point rep
resents a real number.
 number theory
the study of integers.
 numerator
;  oblique angle
: an angle that is not 90
0
I
:  oblique coordinates
I a coordinate system in which
I the axes are not perpendicular.
f,p
in the fraction x/y, x is called the
numerator and y is called the I
denominator.
5
• numerical analysis
the study of methods for ap
proximation of solutions of
various classes of mathematical
problems including error analy
SIS.
 oblate spheroid
an ellipsoid produced by rotat
ing an ellipse through 360
0
about its minor axis.
R
~  oblique line
; a line that has a definite slope
: not equal to zero
I
:  oblique prism or cylinder
~ a nonright prism or cylinder
~  oblique triangle
; a triangle that is not a right tri
angle.
MR.thematics==================== II
112
=========*
• observational study
c.f. controlled experiment.
• obtuse angle
I • odd number
I an integer that is not diVisible
by 2.
an angle whose measure is • odds
greater than 90 but less than I the odds in favour of an event
180 degrees. I is the ratio of the probability
• obtuse triangle
a triangle that contains an ob
tuse angle.
• octagon
an eightsided polygon.
• octahedron
a polyhedron with 8 faces.
that the event occurs to the
probability that the event does
I not occur. For example, sup
pose an experiment can result
in any of n possible outcomes,
I all equally likely, and that k of
the outcomes result in a "win"
and nk result in a "loss."
I Then the chance of winning is
kin; the chance of not winning
is (nk)/n; and the odds in
I favour of winning are (k/n)/
«nk)/n) = k/(nk), which is
the number of favourable out
I comes divided by the number
I of unfavourable outcomes.
~   .  ..
Note that odds are not syn
• octant onymous with probability, but
anyone of the 8 portions of ; the two can be converted back
space dtermined by the 3 coor and forth. If the odds in
dinate planes. favour of an event are q, then
• odd function I the probability of the event is
a function f(x) is called an odd I q/(l +q). If the probability of
function if f(x) =f( x) for all x. an event is p, the odds in
• odd node I favour of the event are p/(l
a node with an odd number of ~ p) and the odds against the
event are (lp)/p.
arcs
II =======MllthmuJnes
II oddsmtio I orderofopemtions
113
*================
 odds ratio ~  open interval
an alternative characterisation ; an interval that does not include
of the parameter 'p' for a bi : its two endpoints.
nomial process is the ratio of ~ "t C
h
. . f .  °PPOSI e laces
t e InCIdences 0 the two al I f: th li' all I I
ternatives : p/( Ip) ; this : aces at e m par e panes
I "
quantity is termed the odds .  opposIte rays
ratio; the value may range ; two rays with a common end
from zero to infinity. This re : point that form a line
I
lates to a possible view of a : _ opposite side
binomial process as the com  ~ (of an angle of a triangle) The
bined activity of two Poisson I side that is not a side of the
processes with a limit upon : angle.
total count for the two pro I
cesses combined. :  OR disjunction, logical
 one to one
a function f is said to be one to
one if f(x) =f(y) implies that
x=y.
 onedimensional
having length, but no width;
examples: a line, a ray, a seg
ment
 onepoint perspective
a method of perspective draw
ing that uses one vanishing
point.
onto
a function f is said to map A
onto B if for every b in B,
there is some a in A such
f(a) =b.
~ disjunction
I an operation on two logical
~ propositions. If p and q are two
propositions, (p OR) q is a
~ proposition that is true if p is
; true or if q is true (or both);
: otherwise, it is false. That is, (p
~ OR) is true unless both p and q
; are false. C.f. exclusive disjunc
: tion, XOR.
I
:  order
~ to place numbers in order from
; smallest to largest or largest to
: smallest.
I
:  order of operations
~ the rule for using operations on
I numbers; first parentheses,
: then exponents, then multipli
MIIthem4.tics======= II
114
=========*
orderedpair lordinatiml II
cation and division, then addi
tion and subtraction.
• ordered pair
I • ordinal variable
a variable whose possible val
ues have a natural order, such
the two numbers that (called
coordinates) are used to iden I
tify a point in a plane; written I
I as {short, medium, long},
{cold, warm, hot}, or {O, 1,
2, 3, ... }. In contrast, a vari
able whose possible valufs are
(x, y)
• ordered pair rule
a rule that uses ordered pairs
to describe a transformation.
For example, the ordered pair
rule (x, y) ??(x + h, y + k) de
scribes a translation horiwn
tally by h units and vertically by
k units.
• ordered triple
I {straight, curly} or {Arizona,
California, Montana, New
York} would not naturally be
I ordinal. Arithmetic with the
possible values of an ordinal
variable does not necessarily
I make sense, but it does make
sense to say that one possible
value is larger than another.
• ordinate
I the ycoordinate of a point in
I the plane.
• ordination
the three numbers (called coor
dinates) that are used to iden
tify a point in space; written (x,
y, z)
a representation of objects with
• ordinal scale I respect to one or more coordi
a measurement type for which I nate axes. There are many kinds
the relative values of data are of ordinations depending upon
defined solely in terms of be I the goals of the ordination and
ing lesser, equalto or greater I criteria used. For example, plot
as compared with other data ting objects according to their
on the ordinal scale. These scores on the first two princi
characteristics may arise from I pal component axes provides
categorical rating scales, or the twodimensional ordination
from converting interval scale best summarising the total vari
data to become ranked data. I ability of the objects in the
original sample space. Biplots
combine an ordination of speci
II =======MsJtbmuJtics
II orientation I orthographic drawing .. = = = = = = = = = = = 1 ~ 1 ~ 5
mens and an ordination ofvari ~ • orthogonal
ables.
• orientation
in an image change, the direc
tion in which the points
named go (i.e., how Ns posi I
tion relates to B's and B's re
lates to C's); either clockwise
or counterclockwise for fig
I at right angles. In linear alge
bra, being "at right angles" is
I defined relative to a symmetric
I matrix P, such as the bending
energy matrix; two vectors x
and yare orthogonal with re
I spect to P if xtpy=O. Principal
warps are orthogonal with re
~ spect to bending energy, and
~ relative warps are orthogonal
with respect to both bending
energy and the sample covari
ures
• origin
the point in a coordinate plane
with coordinates (0,0).
I ance matrix .
• orthic triangle I • orthogonal superimposi
the triangle whose vertices are tion
the feet of the altitudes of a a superimposition using only
given triangle. I transformations that are all
• orthocenter
the point of concurrency of a
triangle's three altitudes (or of
the lines containing the alti
tudes).
A
B
c
I Euclidean similarities, i. e., in
volve only translation, rota
I tion, scaling, and, possibly, re
I flection.
I • orthographic drawing
a drawing of the top, front, and
right side views of a solid that
I preserves their sizes and shapes.
Orthomeans "straight;" the
views of an orthographic draw
I ing show the faces of a solid as
if you were viewing them
"headon."
MMhemlJtics===================== II
116 .. orthommnal I operlapping triangles II
• orthonormal I fied, and the outlier is deter
a set of vectors is orthonormal I mined to be spurious. Other
if each has length unity and all wise, discarding outliers can
pairs are orthogonal with re I cause one to underestimate
spect to some relevant matrix, I the true variability of the mea
P, such as the identity matrix. A surement process.
matrix is orthogonal if its rows
( columns) are orthonormal as I
• overlapping angles prop
erty
a set of vectors.
• outcome space
the outcome space is the set of ~
all possible outcomes of a given I
random experiment. The out
come space is often denoted by
the ~ a p i t a l letter S.
I the property that, if two
angles have the same vertex
and overlap so that the
nonoverlapping parts of the
angles are congruent, then the
angles are congruent.
• overlapping segments
property
• outcome value
the value of the test statistic
for the data as initially ob
served, before any re
randomisation . .
• outgoing angle
the angle formed between the
path of a rebounding object (a
billiard ball, a light ray) and the
surface it collides with (a cush
ion, a mirror).
• outlier
the property that, if two line
segments on the same line
I overlap so that the
I nonoverlapping parts are con
gruent, then the line seg
ments are congruent.
I • overlapping triangles
I triangles that share a side or
angle
an outlier is an observation I
that is many SD's from the
mean. It is sometimes tempt
ing tQ discard outliers, but this I
is imprudent unless the cause
of the outlier can be identi
II =======MRrh_tics
II palindrome I parameter
• palindrome
a positive integer whose digits
read the same forward and
backwards.
• palindromic
a positive integer is said to be
palindromic with respect to a
base b if its representation in
base b reads the same from left
to right as from right to left.
• pandiagonal magic square
a magic square in which all the
broken diagonals as well as the
main diagonals add up to the
magic constant.
• pandigital
117
*================
~ • parallel
; (lines, rays, or line segments),
: lying in the same plane and not
I . .
: mtersectmg.
~ • parallel lines
; two or more coplanar lines
: that have no points in com
~ mon or are identical (eg, the
I same line)
I • parallel planes
~ planes that have no points in
: common
~ • parallelepiped
I a prism whose bases are
paraIleograms.
a decimal integer is called I
pandigital if it contains each of
the digits· from 0 to 9.
• paraboloid
a paraboloid of revolution is a I
surface of revolution produced by ; • parallelogram
rotating a parabola about its axis. : a quadrilateral in which both
~ pairs of opposite sides are par
I aIle!.
I • parameter
: in general, a parameter is a
~ number (an integer, a deci
I mal) indexing a function. For
: instance, the Fdistribution
~ used to test decompositions
I of variance has two par am
MRt'nu======= II
118
========*
parameter I partial warp srores !I
eters, both integers: the I • partial least squares
counts of the degrees of free I partial Least Squares is a mul
dom for the two variances tivariate statistical method for
whose ratio is being tested. In I assessing relationships among
morphometrics, there are I two or more sets of variables
four main kinds of param measured on the same enti
eters: nuisance parameters, ties. Partial Least Squares
which must be estimated to I analyses the covariances be
account for differences not of ~ tween the sets of variables
particular scientific interest; rather than optimizing linear
the geometric parameters, I combinations of variables in
such as shape coordinates, in the various sets. Their compu
which landmark shape is ex tations usually do not involve
pressed; statistical param I the inversion of .
I • partial warp scores
eters, such as mean differ
ences or correlations, by
which biological interpreta I
tion is confronted with that I
partial warp scores are the
quantities that characterise
the location of each specimen
data; and another set of geo
metric parameters, such as
partial warp scores or I
Procrustes residuals, in which
the findings of the statistical
analysis are expressed.
I in the space of the partial
warps. They are a rotation of
the Procrustes residuals
I around the Procrustes mean
configuration. For the nonuni
form partial .warps, the coef
I ficients for the rotation are the
principal warps, applied first
to the xcoordinates of the
• parameter
a numerical property of a popu
lation, such as its mean.
• parametric equations
I Procrustes residuals, then to
the ycoordinates and, for
threedimensional data, the z
I coordinates. Coefficients for
the uniform partial warps are
produced by special.
two equations which express I
the coordinates of x and y as
separate functions of a com
mon variable, called the pa I
rameter that is usually time.
II ====================MlJthmuJrics
II partial warps I pentomino
119
*=========
• partial warps
partial warps are an auxiliary
structure for the interpretation
of shape changes and shape
variation in sets of landmarks.
Geometrically, partial warps are
an orthonormal basis for a
space tangent to Kendall's shape I
space. Algebraically, the partial
warps are eigenvectors of the
bending energy matrix that de
scribes the net local informa
tion in a deformation along
each coordinate axis. Except for
the very largestscale partial
warp, the one for uniform
shape change, they have an ap
proximate location and an ap
~ • patterns in fractals recur
; sive rule
: a rule used to fmd terms in num
~ ber sequences using recursion.
Iff'
. • payo matrIX
; a way of representing what each
player in a game wins or loses,
as a function of his and his
~ opponent's strategies.
I • pedal triangle
. the pedal triangle of a point P
with respect to a triangle ABC
I is the triangle whose vertices
are the feet of the perpendicu
lars dropped from P to the
I sides of triangle ABC.
~ • pell number
proximate scale.
• partition
a partition of an event B is a
collection of events {AI, A2,
A3, . . . } such that the events
in the collection are disjoint,
and their union is B (they ex
haust B). That is, AjAk = {}
unless j = k, and B = Al U A2
U A3 U .. . . If the event B is
not specified, it is assumed to
be the entire outcome space S.
• Pascal's triangle
; the nth term in the sequence 0,
: 1, 2, S, 12, ... defmed by the re
~ currence
; Po=O, P,=l, and P
n
=2P
n
.
1
+P
n
.
; 2'
: • pentadecagon
~ a ISsided polygon
~ • pentagon
; a fivesided polygon.
; • pentagonal number
: a number of the form n(3nl)j
~ 2.
a triangular array of binomial ~ • pentomino
coefficients. ; a fivesquare polyomino.
MAthem4rics======= II
120
=================*
percentile I pennutation II
~ ....
,...... .
• • perfect square
I an integer is a perfect square if
it is of the form m
2
where m is
an integer.
• • perimeter of a polygon
• the sum of the lengths of the
sides of the polygon
• percentile • periodic curve
the pth percentile of a list is the • a curve that repeats in a regu
smallest number such that at • lar pattern.
least p% of the numbers in the
list are no larger than it. The •
pth percentile of a random vari  •
able is the smallest number
such that the chance that the •
random variable is no larger •
than it is at least p%. C.f.
quantile.
• perfect cube
an integer is a perfect cube if it •
is of the form m
3
where m is an
integer.
Closed (Periodic) Cubic S5pline
A
B
c:
• permutation
a permutation of a set is an ar
rangement of the elements of
• perfect number
• the set in some order. If the
set has n things in'it, there are
nl different orderings of its
a positive integer that is equal I
to the sum of its proper divi
sors. For example, 28 is per
fect because 28 = 1 + 2 + •
• elements. For the first ele
ment in an ordering, there are
n possible choices, for the sec
4+7+14.
• perfect power
an integer is a perfect power
if it is of the form mn where
m and n are integers and n> 1.
• ond, there remain n1 pos
• sible choices, for the third,
there are n2, etc., and for the
• nth element of the ordering,
• there is a single choice re
mammg. By the fundamental
II ==========MlltbmuJti&s
II perpendicular I Pitman permutation ; = = = = = = = = = = ~ 1 ~ 2 ~ 1
rule of counting, the total
number of sequences is thus
nx(n1)x(n2)x .. . xl.
Similarly, the number of
orderings of length k one can
form from n> = k things is
nx(n1)x(n2)x ... x(n
k+ 1) = n!j(nk)1. This is de
noted np k' the number of per
mutations of n things taken k
at a time. C.f. combinations.
• perpendicular
intersecting at right angles.
• perpendicular bisector
(of a line segment) A line that
divides the line segment into
two congruent parts (bisects
it) and is also perpendicular to
it.
• perpendicular lines
2 segments, rays, or lines that
form a 90 degree angle
• perpendicular planes
planes in which any two inter
secting lines, one in each plane,
form a right angle.
•
OR
Q
P B
N
C
~ • perspective
; feeling of depth
; • perspective drawing
: a technique of representing
~ threedimensional relation
; ships realistically in a draw
: ing, by drawing objects
~ smaller as they recede into
I the distance.
I. pi
: written 7t the ratio C[D where
~ C is the circumference and D is
I the diameter of a circle;
~ 3.14159265359 ...
~ • piecewise function
: a function consisting of two or
~ more equations, defined for
; specified intervals of the inde
: pendent variable.
I
: • Pitman permutation test
~ named after the statistician
; E.J . Pitman who described
: this test, and the PITMAN
~ permutation test, in 1937;
; this is one of the earliest in
: stances of an exact test. An
~ exact rerandomisation test in
I which the test statistic is the
: difference of means of two
~ samples of univariate inter
I valscale data.
MIIthematics================== II
122
pixel I point of averages !I
=================*
• pixel I thickness and is therefore con
small dot of color that makes I sidered twodimensional.
up computer' and TV screens
• placebo effect
• plane figure
a set bf pO'ints
I plane
that are on a
I • plane geometry
the belief or knowledge that one
is being treated can itself have
an effect that confounds with
the real effect of the treatment.
Subjects given a placebo as a I
I the study of twodimensional
figures in a plane
painkiller report statistically • plane section
significant reductions in pain in I the intersection of a figure with
randomised experiments that I a plane
compare them with subjects I • point
who receive no treatment at all. an undefined term in most de
This very real psychological ef I ductive systems. It has no size,
fect of a placebo, which has no I only location and is therefore
direct biochemical effect, is considered zerodimensional.
called the placebo effect. Ad I You can think of geometric fig
ministering a placebo to the I ures as sets of points. In a Car
control group is thus important tesian coordinate system, a
in experiments with human point's location is represented
subjects; this is the essence of a I by a pair of numbers (x
J
y).
blind experiment.
I • point of averages
• placebo in a scatterplot, the point
a "dummy" treatment that has whose coordinates are the
no pharmacological effect; e.g., I arithmetic means of the cor
a sugar pill. I responding variables. For ex
• plane ample, if the variable X is
an undefined term in most de I plotted on the horizontal axis
ductive systems. A flat s u r f a ~ e I and the variable Y is plotted
that extends infinitely. A plane on the vertical axis, the point
has length and width but no of averages has coordinates
I (mean of X, mean of Y).
II =======Mnthmultics
II pointofconcurrency I polyhedron
• point of concurrency
the point at which more than
two concurrent lines, line seg
ments, or rays intersect.
• point of tangency
123
*=========
I improve any prediction of the
I position (time) of the next event
(of a circle) The single point I
where a tangent line touches a
by reference to the detail of any
I number of preceding observa
I tions. The corresponding distri
bution of intervals between
events is an exponential distri
I bution. The conventional ex
ample of a Poisson processes is
concerned with occurence of
circle.
• Poisson distribution
the Poisson distribution is a dis
crete probability distribution
that depends on one paranleter,
m. If X is a random variable
with the Poisson distribution
I radioactive emissions in a sub
stantial sample of radioactive
with a halflife very much
I longer than the total observa
tion period.
• polarity of a variable
the positivity or negativity of a
I variable; its direction
with parameter m, then the
probability that X = k is Em I
x mkjk!, k = 0, 1,2, ... , where
E is the base of the naruralloga
ritlun and ! is the factorial func
tion. For all other values of k,
the probability is zero. The ex
pected value the Poisson distri
bution with parameter m is 111,
and the standard error of the
Poisson distribution with pa
rameter m is m 1/2 .
I • polygon
I a closed planar geometric fig
ure consisting of line segments
I (tlle sides), each of which inter
I sects exactly two others at end
points forming the polygon'S
angles. Each point of intersec
I tion is a vertex of the polygon.
• Poisson process I • polygonal region
a process whereby t;vents occur the union of a polygon and its
independently in some COI1 lntenor
tinuwn (in many applications, I • polyhedron
time), such that the overall den I a solid whose surface consists
sity (rate) is statistically con of polygons and tlleir interiors,
stant but that it is impossible to I each of which is a face. A line
Mathematics=====================
II
124 P01,;,i1W I population stand4rd deviation II
segment where two faces inter I of a box of numb.ered tickets
sect is an edge. A point of in I is the mean of the list com
tersection of three or more prised of all the numbers on
edges is a vertex. I all the tickets. The population
I mean is a parameter. C.f.
 polyomino :
f
sample mean.
a planar figure consisting 0 I
congruent squares joined edge :  population percentage
toedge. the percentage of units in a
I population that possess a speci
fied property. For example, the
percentage of a given collection
I of registered voters who are
registered as Republicans. If
each unit that possesses the
I property is labelled with "1,"
and each lmit that does not pos
sess the property is labelled
 population I with "0," the population per
a definable set of individual I centage is the same as the mean
units to which the findings of that list of zeros and ones;
from statistical examination of I that is, the population percent
a sample subset are intended I age is the population mean for
to be applied. The population a population of zeros and ones.
will generally much outnum The population percentage is a
ber the sample. In re I parameter. c.f. sample percent
randomisation statistics the
age.
process of applying inferences
based upon the sample to the
population is essentially infor
mal.
 population mean
 population standard
deviation
I the standard deviation of the
values of a variable for a popu
lation. This is a parameter, not
a statistic. C.f. sample standard
deviation.
the mean of the numbers in a I
numerical' population. For ex
ample, the population mean
II =======MJJt __ #U
II populatum I preshape space
• population
a collection of units being stud
ied. Units can be people, places,
objects, epochs, drugs, proce
dures, or many other things.
Much of statistics is concerned
125
*=================
I provement of performance of
I a statistical test.
I • practical number
a practical number is a positive
I integer m such that every natu
I ral munber n not exceeding m
is a sum of distinct divisors of
with estimating numerical
properties (parameters) of an I
entire population from a ran m.
dom sample of units from the • precision
population. I the closeness of repeated mea
surements to the same value.
• preform space
• post hoc ergo propter hoc I
after this, therefore because of
this. A fallacy of logic known
since classical times: inferring
the space corresponding to cen
I tered objects, i. e., differences
in location have been removed.
It is of k(PI) dimensions.
a causal relation from correla
tion. Don't do this at home!
• postulates
premises in a deductive system
accepted without proof.
• power
this is the probability that a
• preimage
I the original object that is re
flected
I • premises
statistical test will detect a
defined pattern in data and
declare the extent of the pat
tern as showing statistical sig
nificance. power is related to
type2 error by the simple for
mula: power = (Ibeta) ; the
motive for this redefinition is I
(in a deductive s y ~ t e m ) State
I ments (including undefined
I terms, definitions, properties of
I algebra and equality, postulates,
and theorems) used to prove
I further conclusions.
I • preshape space
the space corresponding to fig
ures that have been centered
I and scaled but not rotated to
so that an increase in value for
power shall represent im
alignment. It is of k(Pl)1 di
mensions.
Mathematics=====================
II
""12""6==========,,,,p"=·nu;aCie I p1incipal components atlalysis II
• prima facie
latin for "at tirst glance." "On
the face of it. " Prima facie evi
dence for something is infor
mation that at first glance sup
ports the conclusion. On closer
examination, that might not be
true; there could be another
explanation for the evidence.
• prime
a prime number is an integer
I its effect on a circle or sphere.
I An affine transformation takes
circles into ellipses. The princi
I pal axes of the shape change are
I the directions of the diameters
of the circle that are mapped
into the major and minor axes
I of the ellipse. The principal
strains of the change are the
ratios of the lengths of the axes
I of the ellipse to the diameter of
the circle. In the case of the tet
rahedron, there are three prin
larger than 1 whose only posi I
tive divisors are 1 and itself.
I cipal axes, the axes of the ellip
• prime factorisation soid into which a sphen: is de
the unique set of factors of a formed. One has the greatest
number, all of which are prime I principal strain (ratio of axis
numbers. I length to diameter of sphere),
• primitive pythagorean
triangle
a right triangle whose sides are
relatively prime integers.
one the least, and there is a third
perpendicular to both, having
I an intermediate principal strain.
I • principal components
• primitive root of unity
analysis
the eigenanalysis of the
sample covariance matrix.
Principal components (PC's)
can be defined as the set of
the complex number z is a
primitive nth root of unity if ~
Zll = 1 but Zk is not equal to 1 for ;
any positive integer k less than
I vectors that are orthogonal
n.
• principal axes and strains
a change of one triangle into I
another, or of one tetrahedron
into another, can be modelled
as an affine transformation I
which can be parameterised by
both with respect to the iden
tity matrix and the sample
covariance matrix. They can
also be defined sequentially:
the first is the linear combi
nation with the largest vari
II ====================MiJthemllhcs
II principal wa1'"Ps I prism.. 12
7
ance of all those with coeffi I set of p 2D (p4 for
cients summing in square to I 3D data) they form a finite se
I; the second has the largest ries. Together with the uniform
variance (when normalised I terms, the partial warps, which
that way) of all that are , are projections (shadows) of the
uncorrelated with the first principal warps, supply an or
one; etc. One way to compute thonormal basis for a space that
principal components is to use I is tangent to Kendall's shape
a singular value decomposi space in the vicinity of a mean
tion. Relative warps are prin form.
• prism
cipal components of partial I
warp scores. There is a lot to
be said about PC's; see any of
the colored books.
I a polyhedron with two congru
ent polygons in parallel planes
as bases. Line segments (lateral
I edges) connect the correspond • principal warps
ing bases to form lateral faces,
which are parallelograms. An
I altitude is a line segment be
I tween, and perpendicular to, the
planes of the bases. The height
I is the length of an altitude. If
I the lateral edges are perpen
dicular to the bases, the prism
is a right prism; otherwise it is
I oblique.
principal warps are
eigenfunctions of the bending
energy matrix interpreted as
actual warped surfaces (thin
plate splines) over the picture
of the original landmark con
figuration. Principal warps are
like the harmonics in a Fourier
analysis (for circular shape) or
Legendre polynomials (for lin
ear shape) in that together they
decompose the relation of any
sample shape to the sample av ,
erage shape as a unique swu
mation of multiples of
eigenfunctions of bending en ,
ergy. They differ from these
more familiar analogues in that
there are only p3 of them for a I
MRthemilties=================== II
""12,,,,8========== .probability I probabilitydistrilmtilm "
• probability I have probability density func
the probability of an event is a I nons.
I • probability distribution
number between zero and
100%. The meaning (interpre
tation) of probability is the sub
ject of theories of probability,
which differ in their interpreta
tions. However, any rule for I
assigning probabilities to events
has to satisfy the axioms of ~
probability.
• probability density func
tion
the chance that a continuous
the probability distribution of
I a random variable specifies
I the chance that the variable
takes a value in any subset of
the real numbers . (The sub
sets have to satisfy some tech
nical conditions that are not
important for this course.)
I The probability distribution of
a random variable is com
pletely characterised by the
cumulative probability distri
bution function; the terms
sometimes are used synony
mously. The probability distri
bution of a discrete random
random variable is in any I
range of values can be calcu
lated as the area under a curve
over that range of values. The I
curve is the probability den I
sity function of the random
variable. That is, if X is a con
tinuous random variable,
there is a function f(x) such
that for every pair of numbers
a<=b, P(a<= X <=b) =
(area under f between a and
b); f is the probability density
function of X. For example,
the probability density func
tion of a random variable with
variable can be characterised
I by the chance that the random
I variable takes each of its pos
sible values. For example, the
probability distribution of the
I total number of spots S show
ing on the roll of two fair dice
can be written as a table:
I s P(S=s)
2 1/36
3 2/36
a standard normal distribu
tion is the normal curve. Only
continuous random variables
I 4 3/36
5 4/36
6 5/36
I 7 6/36
II ==;;;;;;=======Mat"tics
II probability wmts I probability} 0""if==========1""2,,,,9
8 5/36 • probability sarr.ple
9 4/36 I a sample drawn from a popula
10 3/36 : tion using a random mecha
11 2/36 nism so that every element of
12 1/36 ; the population has a known
The probability distribution of : chance of ending up in the
a continuous random variable sample.
can be characterised by its prob
• probability, theories of
ability density ftillction.
I a theory of probability is a way
• probability events : of assigning meaning to prob
the set of all possible outcomes ability statements such as "the
of an experiment is the sample I chance that a thumbtack lands
space; any subset of the sample pointup is 2/3." That is, a
space IS an event. theory of probability connects
• probability histogram I the mathematics of probabil
a probability histogram for a I ity, which is the set of conse
d bl I
q uences of the axioms of
ran om varia e is ana ogous
probability, with the real
to a histogram of data, but in
; world of observation and ex
stead of plotting the area of
the bins proportional to the periment. There are several
.. common theories of probabil
relative frequency of observa
tions in the class interval, one I ity. According to the fre
plots the area of the bins pro ; quency theory of probability;
portional to the probability the probability of an event is
that the random variable is in I the limit of the percentage of
times that the event occurs in
the class interval.
0.45
0.4
.
!1 0.3
f
L. 0.25
0.2
0.15
0.1
0.05
o
1
_oI.go,x
repeated, independent trials
under essentially the same cir
; cumstances. According to the
: subjective theory of
ity; a probability is a number
; that measures how strongly
: we believe an event will occur.
The number is on a scale of
Mslt'tics======= "
=13=0========'1'== ;ocrustei distance I procrustes methods 11
0% to 100%, with 0% indicat I most subsequent morphomet
ing that we are completely I ric analyses .
I • procrustes methods
sure it won't occur, and 100%
indicating that we are com
pletely sure that it will occur. I
According to the theory of :
equally likely outcomes, if an ~
experiment has n possible out I
comes, ana (for example, by
symmetry) there is no reason
that any of the n possible out I
comes should occur preferen
tially to any of the others, then
the chance of each outcome is I
100%jn. Each of these theo
ries has its limitations, its pro
ponents, and its detractors.
a term for leastsquares meth
ods for estimating nuisance
parameters of the Euclidean
similarity transformations.
The adjective "Procrustes"
I refers to the Greek giant who
would stretch or shorten vic
tims to fit a bed and was first
I used in the context of super
imposition methods by Hudey
and Cattell, 1962, The
I Procrustes program: produc
ing a direct rotation to test an
hypothesised factor structure,
• procrustes distance I Behav. Sci. 7:258262. Mod
approximately ,the square root I ern workers have often cited ·
of the sum of squared differ : Mosier (1939), a psychome
ences benveen the positions of ~ trician, as the earliest known
the landmarks in two optimally; developer of these methods.
(by leastsquares) superim However, Cole (1996) reports
posed configurations at centroid that Franz Boas in 1905 sug
size. This is the distance that I gested the "method of least
defines the metric for Kendall's differences" (ordinary
shape space. Procrustes analysis) as a
I means of comparing homolo
• procrustes mean
the shape that has the least I gous points to address obvi
ous problems with the stan
summed squared Procrustes
distance to all the configurations I dard pointline registrations
of a sample; the best choice of ~ (Boas, 1905). Cole further
points out that one of Boas'
consensus configuration for ; students extended the method
11 =======MiJthmulries
II procrustes residttals I proportional
131
*=================
to the construction of mean I • procrustes superimposi
configurations from the super tion
imposition of multiple speci the construction of a twoform
mens using either the stan I superimpOSitIOn by least
dard registrations of Boas' I squares using orthogonal or af
method (Phelps, 1932). The fine transformations.
latter being essentially a Gen
eralized Procrustes Analysis. I
• pronic number
a number of the form n(n+ 1) .
• procrustes residuals
• proof
I a sequence of justified conclu
the set of vectors connecting the
landmarks of a specimen to cor
responding landmarks in the I
consensus configuration after a
Procrustes fit. The sum of
squared lengths of these vectors
is approximately the squared
Procrustes distance between the
sions used to prove the validity
of an ifthen statement
• proper divisor
I the integer d is a proper divisor
of the integer n if O<d<n and
d is a divisor of n.
specimen and the consensus in I • proper subset
Kendall's shape space. The par I a subset that doesn't include
tial warp scores are an orthogo everything in its parent set.
nal rotation of the full set of I
these residuals.
• procrustes scatter
a collection of forms all super
imposed by ordinary orthogo
nal Procrustes fit over one
single consensus configura
tion that is their Procrustes
mean; a scatter of all the
Procrustes residuals each cen
tered at the corresponding
landmark of the Procrustes
mean shape.
• properties of equality
reflexive property (a = a),
I symmetric property (if a = b,
then b = a), and transitive
property (if a = band b = c,
I then a = c).
• proportion
I a statement of equality between
two ratios.
• proportional
lone of four numbers that form
I a true proportion
MRthem4tics=================== II
132
= = = = = = = ~ .
proportWnality I ppalue II
• proportionality
a relationship described by a
constant ratio.
• proposition, logical
proposition
I process. Successive pseudoran
I dom data are produced by a
fLxed calculation process acting
upon preceding data from the
I pseudorandom sequence. To
start the sequence it is necessary
to decide arbitrarily upon a first
I datum, which is termed the
seed value ..
a logical proposition is a state I
ment that can be either true or
false. For example, "the sun is
shining in Berkeley right now" I
is a proposition. • pvalue
• protractor
a tool used to measure the size
of an angle if'. degrees.
• prove
use logical arguments, defini
tions, theorems, and properties
to show that a relationship is
true for all numbers or specific
set of figures.
• pseudorandom
suppose we have a family of
I hypothesis tests of a null hy
pothesis that let us test the hy
pothesis at any significance level
I p between 0 and 100% we
choose. The P value of the null
hypothesis given the data is the
I smallest significance level p for
I which any of the tests would
have rejected the null hypoth
esis. For example, let X be a test
I statistic, and for p between 0
and 100%, let xp be the small
est number such that, under the
I null hypothesis, P( X < = x )
> = p. Then for any p between
o and 100%, the rule reject the
I null hypothesis if X < xp tests
the null hypothesis at signifi
cance level p. If we observed X
I = x, the Pvalue of the null hy
a source of data which is effec I
tively unpredictable although I
generated by a determinate
pothesis given the data would
be the smallest p such that x <
xp.
II ==========MRth_ties
II pyramid I qef
• pyramid
a polyhedron with a polygon
base and line segments con
necting the vertices of the
base with a single point (the
vertex of the pyramid) that is
not coplanar with the base.
The altitude is the line seg
ment from the vertex ending
at and perpendicular to the
plane of the base. The height
is the length of the altitude.
If the line segment connecting
the vertex to the center of the
base is perpendicular to the
base, then the pyramid is
right; otherwise it is oblique.
• pythagorean fractal similar
figures
figures that have the same
shape but not necessarily the
same size. Their corresponding
sides are proportional.
• pythagorean theorem
the theorem that says that, in a
right triangle, the sum of the
133
. = = = = = = = = = = = = = = ~
I squares of the lengths of the
; legs equals the square of the
: length of the hypotenuse.
I
: • pythagorean triangle
~ a right triangle whose sides are
I integers.
; • pythagorean triple
: three positive integers with
I
: the property that the sum of
I the squares of two of the in
: tegers equals the square of the
third. If the three integers
I have no common integer fac
I tors, then the triple is primi
: tive. If the three integers have
I a common factor, then the
; triple is a multiple.
; • qed
: ((quod erat demonstrandum))
~ (Latin) This stems from medi
; eval translators' habitual ten
: dency of translating the Greek
~ for "this was to be demon
; strated" to the Latin phrase
: above. This appeared originally
~ at the end of many of Euclid's
I propositions, signifying that he
~ had proved what he set out to
: prove.
I
: .qef
I ((quod erat faciendum)) is the
~ latin for "which was to be
: done" It appears in Latin
MAthmuJti&s======= II
134 .. qw:ulrangle I quantitative l1ariable 1/
translations of Euclid's works I • quadrature
signifying that he had demon I the quadrature of a geometric
strated what he had set out to figure is the determination of
demonstrated. I its area.
• quadrangle
a closed broken line in the plane
consisting of 4 line segments.
• quadrangular prism
a prism whose base is a quadri
lateral.
• quadrangular pyramid
a pyranlid whose base is a quad
rilateral.
I • quadric curve
I the graph of a second degree
equation in two variables.
• quadric surface
the graph of a second degree
I equation in three variables.
I • quadrilateral
a foursided polygon ..
• quadrinomial
an algebraic expression consist
ing of 4 terms.
• quadrant :
anyone of the four portions of I
the plane into wruch the plane :
I • qualitative variable
is divided by the coordinate a qualitative variable is one
axes. whose values are adjectives,
• quadratfrie I such as colours, genders, na
square free tionalities, etc. C.f. quantitative
variable . and categorical vari
• quadratic equation I able.
an equation of the form f(x) =0
where f(x) is a second degree I • quantitative variable
polynomial. That IS, I a variable that takes numerical
ax
2
+bx+c=0. values for which arithmetic
• quadratic term of an
equation
I makes sense, for example,
I counts, temperatures, weights,
amounts of money, etc. For
some variables that take nu
the term AX2 in a quadratic I
equation
I merical values, arithmetic with
those values does not make
sense; such variables are not
II =======M1It __ ries
IllJUllrtic polynomial I radical axis
quantitative. For example, add
ing and subtracting social secu
rity numbers does not make
sense. Quantitative variables
typically have units of measure
ment, such as inches, people, or
pounds.
 quartic polynomial
a polynomial of degree 4.
135
*================
~ est integer that is at least as
; big as the number of entries
: in the list divided by two. Call
~ that integer 1. The lth element
; of the sorted list is the me
: dian. Find the smallest integer
~ that is at least as large as the
1 number of entries in the list
~ times 3/4. Call that integer m.
: The mth element of the sorted
 quartiles 1 list is the upper quartile.
there are three quartiles. The ~ Quota Sampling.
first or lower quartile (LQ) of : .. I 'al
. .  qUlnnc po ynoml
a hst IS a number (not neces 1 .
sarily a number in the list) ~ a polynoffilal of degree 5.
such that at least 1/4 of the :  quotient
numbers in the list are no 1 the result of a division.
larger than it, and at least 3/4 ; _ R
of the numbers in the list are
no smaller than it. The second
I
· rotation
quartile is the median. The ;  r
third or upper quartile (UQ) : radius
is a number such that at least ~ _ radian
3/4 of the entries in the list are 1 a unit of angular measurement
no larger than it, and at least such that there are 2 pi radians
1/4 of the numbers in the list in a complete circle. One radian
are no smaller than it. To find ~ = ISOjpi degrees. One radian
the quartiles, first sort the list ; is approximately 57.30.
into increasing order. Find the
smallest integer that is at least
as big as the number of entries
in the list divided by four. Call
that integer k. The kth ele
ment of the sorted list is the
;  radical axis
: the locus of points of equal
1 • h
: power Wit respect to two
1 circle.
lower quartile. Find the small
.MR.thmulties=================== II
136
========*
radical center I random sample II

0
A '"',\.
 f \0'" B
. '\.
\
J '
. \
J \
, \
,
\.
I each time a measurement is
I made, and behaves like a num
ber drawn with replacement
I from a box of numbered tick
I ets whose average is zero.
I • random experiment
an experiment or trial whose
outcome is not perfectly pre
I dictable, but for which the long
• radical center run relative frequency of out
the radical center of three circles I comes of different types in re
is the common point of; peated trials is predictable. Note
interesection of the radical axes that "random" is different from
of each pair of circles. "haphazard," which does not
• radii
plural form of radius
• radius
the segment whose endpoints
are any point on a circle or
sphere and its center; the length
of that segment
• radix point
I necessarily imply longterm
regularity.
• random sample
a random sample is a sample
I whose members are chosen at
I random from a given popula
tion in such a way that the
I chance of obtaining any par
I ticular sample can be com
puted. The number of units in
the sample is called the sample
I size, often denoted n. The
the generalisation of decimal
point to bases of numeration I
other than base 10.
• random error
all measurements are subject to
error, which can often be bro
ken down into two components:
a bias or systematic error, which
affects all measurements the
same way; and a random error,
which is in general different
number of units in the popu
lation often is denoted N.
I Random samples can be
drawn with or without replac
ing objects between draws;
I that is, drawing all n objects
in the sample at once (a ran
dom sample without replace
II ===================MJJt __ ncs
111
o
llndomVilriable I mndomislltiunset .. 137
ment), or drawing the objects ~ 1 to the outcome {T, T, H}, the
one at a time, replacing them ; number to to the outcome {T:
in the population between : H, H}, and the number 3 to the
draws (a random sample with ~ outcome {H, H, H}.
replacement). In a random ~ _ randomisation
sample with replacement, any ; the process of arranging for
given member of the popula : datacollection, in accordance
tion can occur in the sample ~ with the experimental design,
more than once. In a random I such that there should be no
sample without replacement, : foreseeable possibilty of any
any given member of the ~ systematic relationship be
population can be in the
~ tween the data and any
sample at most once. A ran
; measureable characteristic of
dom sample without replace : the procedure by which the
ment in which every subset of I did h" .
f h N
. . h ata was samp e . t IS IS usu
not e umts m t e popu : 11 d bOO
1
0 0 11 1
0
k 1 0 1 I a Y arrange Y asslgnmg ex
anon IS equa y 1 e y IS a so . I'
11 dOl d : penmenta umts to groups,
ca e 1 a
Th
slm
p
e random ~ and repeated measures to ex
samp e. e term ran om· 0 l' 0 I
1 0 hid I pen menta umts, on a stnct y
samp e WIt rep acement e: random basis.
notes a random sample drawn I
in such a way that every n :  randomisation distribu
tuple of units in the popula ~ tion
tion is equally likely. ; a collection of values of the test
 random variable
a random variable is an assign
ment of numbers to possible
outcomes of a random experi
ment. For example, consider
tossing three coins. The num
ber of heads showing when the
coins land is a random variable:
it assigns the number 0 to the
outcome {T, T, T}, the number
: statistic obtained by undertak
~ ing a number of re
I randomisations of the actual
~ data within the randomisation
: set.
I d"
:  ran olDlsabon set
~ the collection of possible re
; randomisations of data within
: the constraints of the experi
~ mental design
MIIth_tics================== /I
",1""3,,,,8===========;ndmnisation test I ratio of similitude II
• randomisation test I signed to treatment, and the
the rationale of a randomisation I rest to control. Randomising
test involves exploring re the assignment tends to de
randomisations of the actual I crease confoimding of the treat
data to form the randomisation I ment effect with other factors,
distribution of values of the test by making the treatment and
statistic. the outcome value control groups roughly compa
value of the test statistic is I rable in all respects but the
judged in terms of its relative treatment.
position within the re
randomisation distribution. if I
• range
the range of a set of num bers is
I the largest value in the set mi
nus the smallest value in the set.
the outcome value is near to one
extreme of the re
randomisation distribution then I
it may be judged that it is in the
extreme tail of the distribution,
with reference to a nominal al I
Note that as a statistical term,
I the range is a single number,
not a range of numbers.
• ranked data
pha criterion value, and thus I
judged to show statistical sig
this refers to the practice of tak
I ing a set of N data, to be re
I garded as ordinalscale, and re
nificance.
• randomised controlled
experiment
placing each datum by its rank
I (1 .. N) within the set.
an experiment in which chance
is deliberately introduced in as
signing subjects to the treat
ment and control groups. For
example, we could write an
identifying number for each I
subject on a slip of paper, stir
up the slips of paper, and draw
slips without replacement until
we have drawn half of them.
The subjects identified on the
slips drawn could then be as
• rate
I a ratio where the quantities are .
of different kinds; example: 60
miles per hour
• ratio
I the quotient of two numbers.
I • ratio of similitude
the simplest form ratio of the
measures of corresponding
I parts of similar figures.
II =======MR.them4tics
II ratio scale I recursive form
 ratio scale
this is a type of measurement
scale for which it is meaning
ful to reason in terms of dif
ferences in scores and also in
terms of ratios of scores. Such
a scale will have a zero point
which is meaningful in the
sense that it indicates complete
absence of the property which
the scale measures. The ratio
scale may be either unipolar
(negative values not meaning
ful) or bipolar (both positive
and negative values meaning
ful), and either continuous or
discrete.
 rational number
a rational number is a number
that is the ratio of two integers.
All other real numbers are said
to be irrational.
 ray
all points on a line that lie on
one side of a specified point, the
ray's endpoint. A ray is referred
to by giving the names of two
points, first the endpoint and
then any point on the ray.
139
*================
~  real part
; the real number x is called ther
: eal part of the complex number
~ x+iy where x and y are real and
; i=sqrt( 1).
;  real variable
: a variable whose value ranges
~ over the real numbers.
I • I
:  reclproca
I the reciprocal of the number x
: is the number l/x.
I
;  recompose
: put addends or factors of a
I given number back together in
a way different from the origi
nal arrangement or decompo
I sition.
~  rectangle
; a quadrilateral with 4 right.
: angles.
I
:  rectangular solid
I the union of a box and its inte
I nor
;  recursion
: the process of generating a se
~ quence (or pattern) by speci
; fying a first term and then ap
: plying a rule to obtain any suc
 real axis ~ ceeding term from the previ
the xaxis of an Argand diagram. I ous term.
 real numbers
;  recursive form
rational and irrational numbers.
a formula for the next term of
I rejlectionalsymmetry II
a sequence given the term be
fore it.
• reference angle
I part of the analysis, the bend
I ing energy that goes with
them is computed using the
I geometry of the grand mean
shape, and the orthogonality
that characterises the partial
warps is with respect to this
the angle of less than 360 de
grees that corresponds to an I
angle of over 360 degrees; In I
order to get the reference angle,
you must subtract 360 degrees
from the given angle until there I
is less than 360 degrees left.
I particular formula for bend
ing energy. There has been
some controversery regarding
I the choice of reference .
• reference configuration
I • refine
in the context of superimpo
sition methods, this is the con I
figuration to which data are
to change a conjecture slightly
so that it is true
fit. It may be another speci • reflection
men in the sample but usually I an isometry under which every
it will be the average (consen point and its image are on op
sus) configuration for a posite sides of a fL\:ed line (the
sample. The construction of line of reflection, or mirror
twopoint shape coordinates; line) and are the same distance
does not involve a reference from the line.
specimen, though the intelli I
• reflection image of a figure
gent choice of baseline for the I the set of all of the reflection
construction usually does. The . fi
reference configuration corre
I images of pomts in the Igure
I • reflection notation
rm(ABC), \vhich stands for the
reflection over line m of figure
I ABC
I • reflectional symmetry
sponds to the point of tan
gency of the linear tangent
space used to approximate
Kendall's shape space. The
mean configuration is usually
used as the reference in order
the property of a figure that it
to mInImise distortions I coincides with its image under
caused by this approximation. I at least one reflection. Also
When splines and warps are
II ===================MRther_#&s
II rejlectWnsymmetric M'tre I the mmn, regression effict 141
called line symmetry or mirror • regression fallacy
symmetry. ; the regression fallacy is to at
• reflectionsymmetric
figure
: tribute the regression effect to
I
. an external cause.
a figure that shows reflection I
symmetry
• regression toward the
mean, regression effect
• reflex angle
an angle between 180
0
and
3600.
• reflex polygon
a polygon for which 2 or more
of its sides intersect each other
suppose one measures two vari
I
: abIes for each member of a
I group of individuals, and that
the correlation coefficient of the
: variables is positive (negative).
I If the value of the first variable
for that individual is above av
• reflexive property of erage, the value of the second
congruence I variable for that individual is
the property of every geomet I likely to be above (below) av
ric object that it is congruent to erage, but by fewer standard
itself. I deviations than the first variable
• region
I is. That is, the second observa
tion is likely to be closer to the
mean in standard units. For ex
I ample, suppose one measures
the tullOn of a figure and its in I
terior
• regression
a model for predicting one vari
able from another. Due to
Francis Galton, the word comes
from the faG: that when mea
surements of offspring,
whether peas or people, were
plotted against the same mea
surements of their parents, the
offspring measurements "went
back" or regressed towards the
mean.
the heights of fathers and sons.
Each individual is a (father, son)
I pair; the two variables mea
sured are the height of the fa
ther and the height of the son.
These two variables will tend to
I have a positive correlation co
: efficient: fathers who are taller
than average tend to have sons
; who are taller than average.
: Consider a (father, son) pair
chosen at random from this
MRth_tics======= II
142 regres;, linearngnssWn I regularpyramid II
group. Suppose the father's I zontal axis, the regression line
height is 3SD above the aver I passes through the point of av
age of all the fathers' heights. erages, and has slope equal to
(The SD is the standard devia I the correlation coefficient times
tion of the fathers' heights.) I the SD ofY divided by the SD
Then the son's height is also of X.
• regular hexagon
~ k e l y to be above the average
of the sons' heights, but by I
fewer than 3SD (here the SD is
the standard deviation of the
a sixsided figure whose sides
I are of equal length and whose
angles are of equal measure.
• regular polygon
a convex polygon whose angles
and sides are all congruent
sons' heights). In an hypothesis I
test using a test statistic, the I
rejection region is the set of :
values of the test statistic for I
which we reject the null hypoth··
I • regular polyhedron
eSlS. a polyhedron whose faces are
• regression, linear regres enclosed by congruent, regular
sion I polygons that meet at all verti
linear regression fits a line to a I ces in exactly the same way.
scatterplot in such a way as to
minimise the sum of the I
squares of the residuals. The I
resulting regression line, to
gether with the standard devia
tions of the two variables or I
their correlation coefficient, can
be a reasonable summary of a
scatterplot if the scatterplot is I
roughly footballshaped. In
other cases, it is a poor sum
mary. If we are regressing the I
" variable Y on the variable X, and
if Y is plotted on the vertical
axis and X is plotted on the hori  I
• regular pyramid
a pyramid whose base is a regu
lar polygon and whose vertex
forms a segment with the cen
II =========i;;;;;;;;;;;;=========MR.thcmancs
II reguJartessellatWn I
ter of the polygon perpendicu
lar to its plane
_ regular tessellation
an edgetoedge tessellation in
which tiles are congruent regu
lar polygons.
differences. Relative warps can
; be computed from Procrustes
: residuals or from partial warps.
I
:  remote interior angles
(of the exterior angle of a tri
; angle), the two interior angles
: that do not share a vertex with
 relative power the e:A'terior angle.
a comparison of two or more
statistical tests, for the same  rep digit
experimental design, sample I an integer all of whose digits
size, and nominal alpha crite : are the same.
I
rion value, in terms of the re : _ repeated median
spective values of power. a median of medians. Repeated
_ relative warps I medians are used to estimate
relative warps are principal com some superimposition param
ponents of a distribution of : eters in the resistantfit meth
shapes in a space tangent to ods. For example, the resistant
Kendall's shape space. Theyare ; fit rotation estimate is the me
the axes of the "ellipsoid" oc : dian of the estimates obtained
cupied by the sample of shapes for each landmark, which is, in
in a geometry in which spheres ; tmn, the median of angular dif
are defined by Procrustes dis : ferences between the reference
rance. Each relative warp, as a configuration and the configu
direction of shape change about ; ration being fit of the line seg
the mean form, can be inter : ments defined using that land.:
preted as specifying multiples of mark and the other nl land
one single transformation, a I marks. Repeated medians are
transformation that can often : insensitive to larger subsets of
be usefully drawn out as a thin extremely deviant values than
plate spline. In a relative warps I simple medians.
analysis, the parameter can be .
I _ repeatedmeasures
used to weight shape variation this is a feature of an experi
By the geometric scale of shape : mental design whereby sev
I
MJJthematU:s======= II
144
~ ~ ~ ~ ~ ~ ~ = = *
replications I resitlfuU II
eral observations measured ~ • repunit
on a common scale refer to ; an integer consisting onlyofl's.
the same sampling unit. Iden
tification of the relation of the I • rerandomisation
individual observations to the I the process of generating al
. ternative arrangements of
expenmental design is crucial
I given data which would be
to this definition. Examples :
the measurement of water I consistent "\vith the experi
I
mental design.
evel at a particular site on
• rerandomisation statistics
also known as permutation or
randomisation statistics. These
are the specific area of concern
several systematicallydefined :
occasions; measurement of I
reactiontime of an individual :
using right hand and left hand
I of this present glossary.
separatel.
• replications
I • resampling stats
this is the name of an educa
tional initiative involving the
use of a programming lan
guage, in the form of an inter
preter, allowing the user to
this is a feature of an experi
mental design whereby obser
vations on an experimental I
unit are repeated under the I
same conditions. Identifica
tion of the position of a .par
ticular observation within the
sequence of replications is ir
relevant.
I specify montecarloresampling
I of a set of data and accumula
tion of the randomisation dis
I tribution of a defmed test sta
I tistic .
• representative
patterns in a sample of units I • residual
may reasonably be attributed the deviations of an observed
to the population from which : value or vector of values from
the sample is drawn, only if I some expectation, e.g., the dif
the sample is representative. ~ ferences between a shape and
in practical terms, to ensure its prediction by an allometric
that a sample is representative I regression expressed in any set
almost always means ensuring of shape coordinates.
that it is a random sample.
II =======Msthem4ries
II residual plot I resultant'flC&tor'
145
*================
• residual plot ~ mean is not resistant; the me
a residual plot for a regression ; dian is.
is a plot of the residuals from i • resistantfit superimposi
the regression against the ex tion
planatory variable. ~ superimposition methods that
• residual ; use medianand repeatedme
the difference between a datum : dianbased estimates of fitting
and the value predicted for it by ~ parameters rather than least
a model. In linear regression of I squares estimates. Resistantfit
a variable plotted on the verti : procedures are less sensitive to
cal axis onto a variable plotted ~ subsets of extreme values than
on the horizontal axis, a re I those of comparable least
sidual is the "vertical" distance ~ squares methods. As such, their
from a datum to the line. Re : results may provide a simple
siduals can be positive (if the ~ description of differences in
datum is above the line) or ; shape that are due to changes
negative (if the datum is below : in the positions of just a few
the line). Plots of residuals can ~ landmarks. However, resistant
reveal computational errors in i fit methods lack the welldevel
linear regression, as well as con : oped distributional theory asso
ditions under which linear re ~ ciated with the leastsquares fit
gression is inappropriate, such I ting methods.
as nonlinearity and; • resolution
heteroscedasticity. If linear re : the smallest scale distinguish
gression is performed properly, ~ able by a digitising, imaging, or
the sum of the residuals from
the regression line must be
zero; otherwise, there is a com
putational error somewhere.
• resistant
a statistic is said to be resistant
if corrupting a datum cannot
change the statistic much. The
~ display device.
I • resultant vector
~ the result of combining two vec
: tors. To fmd the resultant vec
I tor, slide the original vectors so
~ that their tails intersect. The
: resultant vector's tail is the com
~ mon tail. Its head is the image
146
=================*
of the head of one of the vec ~ • right cylinder
tors after you translate it along; a cylinder whose direction of
the other vector. Also known as sliding is perpendicular to the
a vector sum. I plane of the base
• review mandala
I • right prism
I a prism whose direction of slid
ing is perpendicular to the plane
of the base
Light Ray
   : ~  I   " " " " "
a circular design arranged in I
concentric arcs.
• rhombus
a parallelogram with four equi
lateral sides
• ridge curve
ridge curves are curves on a sur
face along which the curvature
perpendicular to the curve is a
local maximum. For instance on
a skull, the line of the jaw or
the rim of an orbit.
• right angle
an angle whose measure is 90
degrees
• right cone
• right triangle
I a triangle with a right angle.
: The side opposite the right
~ angle is the hypotenuse. The
I other two sides are the legs.
I • rigid motion
a motion that preserves shape
and size.
• rigid rotation
I an orthogonal transformation
of a real vector space with re
spect to the Euclidean distance
I metric. Such transformations
leave distances between points
and angles between vectors un
I changed. A principal compo
a cone whose axis is perpen
dicular to the plane containing I
its base
II rmserrorofregression I rutatWn
nents analysis represents a rigid
rotation to new orthogonal
axes. A canonical variates
analysis does not.
Y2
 rms error of regression
the rms error of regression is
the rms of the vertical residu
als from the regression line.
For regressing Y on X, the
nns error of regression is
equal to (I r2)!f2xSDY, where
r is the correlation coefficient
between X and Y and SDY is
the standard deviation of the
values of Y.
147
* = = = = = = = = = = = = = = = = ~
~ of the elements in the list. It is
; a measure of the average "size"
: of the elements of the list. To
~ compute the rms of a list, you
; square all the entries, average
: the numbers you get, and take
I
: the squareroot of that average.
I
:  RootMeanSquare Error
I (RMSE)
: the RMSE of an an estimator
~ of a parameter is the square
~ root of the mean squared error
; (MSE) of the estimator. In sym
: boIs, if X is an estimator of the
~ parameter t, then RMSE(X) =
; (E( (Xt)2 ) ) !f2. The RMSE of
: an estimator is a measure of the
~ expected error of the estimator.
; The units of RMSE are the
: same as the units of the estima
~ tor.
 RNG ~  rotation
acronym for Random Number ; in effect, a rotation is a turning
Generator. This is a process : of the plane about a point (the
which uses a arithmetic algo ~ center of rotation) by an angle
rithm to generate seyuences of I (the angle of rotation). For
pseudorandom numbers. : mally, a rotation is an isometry
~ that is the composition of re
I flections through two lines that
~ intersect at the center of the
: rotation. The angle of rotation
~ has twice the measure of the
; smaller angle formed by the
: lines.
 root of unity
a solution of the equation xn = I,
where n is a positive integer.
 rootmeansquare (rms)
the rms of a list is the square
root of the mean of the squares
Mnth_tics======== II
I samplepercentage II
• rotational symmetry
• sample mean
a figure has rotational symme
try if it can be rotated (turned)
less than 360 degrees about a I
point so that it appears the same
as the original figure.
• ruled surface
a surface formed by mqving a
straight line (called the genera I
tor).
• rusty compass
a pair of compasses that are
fixed open in a given position.
• sa
surface area
• sample
a set of individual units, drawn
the arithmetic mean of a ran
dom sample from a popula
tion. It is a statistic commonly
used to estimate the popula
tion mean. Suppose there are
n data, {xl, x2, ... ,xn}. The
sample mean is (xl + x2 + ..
. + xn)jn. The expected value
of the sample mean is the
population mean. For sam
pling with replacement, the
SE of the sample mean is the
population standard devia
tion, divided by the square
root of the sample size. For
sampling without replace
ment, the SE of the sample
mean is the finitepopulation
correction ((Nn)j(Nl»V2
times the SE of the sample
mean for sampling with re
placement, with N the size of
the population and !1 the size
of the sample.
from some definable popula
tion of units, and generally a
small proportion of the popu
lation, to be used for a statis I
tical examination of which the
findings are intended to be
applied to the population. it is I
essential for such inference I • sample percentage
that the sample should be rep the percentage of a random
resentative. in re I sample with a certain prop
randomisation statistics the I erry, such as the percentage of
process of applying inferences voters registered as Demo
based upon the sample to the I crats in a simple random
population is essentially infor I sample of voters. The sample
I
mean is a statistic commonly
rna. I
used to estimate the popula
II ===================Msthematies
II samplesize I sample sum .. 149
tion percentage. The expected ~ • sample standard deviation
value of the sample percent ~ the sample standard deviation
age from a simple random : S is an estimator of the stan
sample or a random sample ~ dard deviation of a population
with replacement is the popu I based on a random sample
lation percentage. The SE of : from the population. The
the sample percentage for ~ sample standard deviation is a
sampling with replacement is I statistic that measures how
(p(lp)/n )V2, where p is the : "spread out" the sample is
population percentage and n ~ around the sample mean. It is
is the sample size. The SE of I quite similar to the standard
the sample percentage for ~ deviation of the sample, but in
sampling without replacement : stead of averaging the squared
is the finitepopulation cor ~ deviations (to get the rms of the
rection ((Nn)/(NI»V2 times ~ deviations of the data from the
the SE of the sample percent : sample mean) it divides the
age for saQlpling with replace ~ sum of the squared deviations
ment, with N the size of the ~ by (number of data 1) before
population and n the size of : taking the squareroot. Suppose
the sample. The SE of the ~ there are n data, {xl, x2, ... ,
sample percentage is often es ; xn}, with mean M = (xl + x2
timated by the bootstrap. : + ... + xn)/n. Then s = ( ((xl
I
• sample size
the number of experimental
units on which observations are
considered. this may be less than
the number of observations in
a dataset, due to the possible
multiplying effects of multiple
variables and/or repeated mea
sures within the experimental
design.
: M)2 + (x2 M)2 + ... + (xn
I M)2)/(nl) )lh The square of
~ the sample standard deviation,
: S2 (the sample variance) is an
I unbiased estimator of the
; square of the SD of the popula
: tion (the variance of the popu
~ lation).
~ • sample sum
; the sum of a random sample
: from a population. The ex
~ pected value of the sample
MnthfR'llBtW======== II
150 .. samplesuwey I samplingerror II
sum is the sample size times I the sampling distribution of an
the population mean. For sam I estimator is the probability dis
pIing with replacement, the tribution of the estimator when
SE of the sample sum is the ~ it is applied to random samples.
population standard devia ; The tool on this page allows you
tion, times the squareroot of : to explore empirically the sam
the sample size. For sampling ~ pling distribution of the sample
without replacement, the SE I mean and the sample percent
of the sample sum is the fi age of random draws with or
nitepopulation correction ~ without replacement draws
«Nn)j(Nl))V2 times the SE I from a box of numbered tick
of the sample sum for sam ~ ets.
pIing with replacement, with : _ lin
N h . f th I· I samp g error
t e Slze .0 e popu a t l ~ n : in estimating from a random
and n the Slze of the samp e. I sample, the difference be
 sample survey tween the estimator and the
a survey based on the responses parameter can be written as
of a sample of individuals, I the sum of two components:
rather than the entire popula I bias and sampling error. The
tion. bias is the average error of the
I estimator over all pos,sible'
I samples. The bias is not ran
dom. Sampling error is the
component of error that var
I ies from sample to sample.
 sample variance
the sample variance is the
square of the sample standard I
deviation S. It is an unbiased
estimator of the square of the
population standard deviation,
which is also called the variance
: The sampling error is ran
~ dom: it comes from "the luck
I of the draw" in which units
of the population.
 sample
a sample is a collection of units
from a population.
 sampling distribution
happen to be in the sample. It
is the chance variation of the
I estimator. The average of the
sampling error over all pos
sible samples (the expected
I value of the sampling error)
II sampling unit I score.. 151
is zero. The standard error of I weight against height, the x
the estimator is a measure of I coordinate of each point would
the typical size of the sampling be height of one person, the
error. I ycoordinate of that point
• sampling unit
I would be the weight of the
same person. In a scatterplot
of height against weight, the
I xcoordinates would be the
a sample from a population I
can be drawn one unit at a
time, or more than one unit
at a time (one can sample clus I
ters of units). The fundamen
tal unit of the sample is called
the sampling unit. It need not
be a unit of the population.
• scale factor
the ratio of corresponding
lengths in similar figures .
• scalene triangle
weights and the ycoordinates
would be the heights .
• scientific notation
I a notation for expressing very
large and very small numbers
as a product of a decimal
I number greater than or equal
to one and less than ten and a
power of ten.
• score
I a linear combination of an ob
a triangle with three sides of
different lengths.
I served set of measured vari
• scatterplot ables. The coefficients for the
a scatterplot is a way to I linear combination are usually
visualise bivariate data. A I determined by some matrix
scatterplot is a plot of pairs of computation. Multivariate
measurements on a collection I statistical findings in the form
of "individuals" (which need I of coefficient vectors can usu
not be people). For example, ally be more easily inter
suppose we record the heights preted if scores are also shown
and weights of a group of 100 I case by case, their scatters,
people. The scatterplot of their loadings (correlations
those data would be 100 with the original variables),
points. Each point represents I
one person's height and
weight. In a scatterplot of
etc.
MRthematics======= II
152
=========*
sd line I selection bias II
• sd line
I • secular trend
a linear association (trend)
with time.
• segment
for a scatterplot, a line that I
goes through the point of av
erages, with slope equal to the I
ratio of the standard devia
tions of the two plotted vari
ables. If the variable plotted
on the horizontal axis is called I
X and the variable plotted on
the vertical axis is called Y, the
slope of the SD line is the SD
of Y, divided by the SD of X.
I aka line segment; the set of
• se(sample mean)
= nV2xSD(box,where
I points consisting of two dis
tinct points and all inbetween
them.
• segment of a circle
I the region between a chord
and the included arc.
SD(box) is the standard devia I
tion of the list of the numbers
on all the tickets in the box (in
cluding repeated values).
• se(sample sum)
= nl/2 x SD(box),and the stan I
dard error of the sample mean
of n random draws with re I
• selection bias
placement from a box of tick I
a systematic tendency for a
I sampling procedure to include
and/or exclude units of a cer
tain type. For example, in a
I quota sample, unconscious
ets is
• secant
a line that intersects a circle
in two points.
• section
(of a solid) An
with a plane.
intersection
• sector of a circle
the region between a central
angle and the arc it intercepts.
prejudices or predilections on
the part of the interviewer can
I result in selection bias. Selec
tion bias is a potential prob
lem whenever a human has
I latitude in selecting individual
II =======M4them4tks
II selfseIeaitm I set
153
*================
 semicircle
i an arc of a circle whose end
: points are the endpoints of a

 semimagic square
i a square array of n numbers
 selfselection : such that sum of the n numbers
selfselection occurs when in in any row or column is a con
dividuals decide for them I stant (known as the magic
selves whether they are in the sum).
control group or the treat . _ 'gu1 t 11 0'
units for the sample; it tends
to be eliminated by probabil
ity sampling schemes in which
the interviewer is told exactly
whom to contact (with no
room for individual choice).
. ". semtre ar esse a on
ment group" SelfselectIon IS I t 11 t" " t' f
" """ a esse a lon conS1S lllg 0
qUlte common III studles of ill all f h
human behaviour. For ex : ar PI? ygons hOw" ose
1 d
" f h ffi f I vertIces le on ot er vertIces,
amp e, stu les 0 tee ect 0 d" h" h "
ki h h alth
" : an tn W lC every vertex IS
smo r ng on "e " . tn I surrounded by the same ar
volve selfselectIon: tndlVldu : t f' I ( t'
I
J: h 1 I range men 0 po ygons 0
a s choose lor t emse ves k' d )" h
whether or not to smoke.
: one or more In s III t e
same order. Also called a 1
Selfselection precludes an ex i uniform tiling.
periment; it results in an ob
servational study. When there
is selfselection, one must be
wary of possible confounding
from factors that influence
individuals' decisions to be
long to the treatment group.
 selfsimilarity
the property of a figure that it
is similar to, or approximately
similar to, a part of itself.
i septagon
: a sevensided polygon
I
:  sequence
a collection of numbers in a pre
; scribed order: a
p
a
4
, •••
iseries
: the sum of a fmite or infinite
I
: sequence
 set
i a set is a collection of things,
: without regard to their order.
I
154
set of dam ptlints I shape lIariable II
• set of data points I • shape coordinates
data collected and placed into I in the past, any system of dis
ordered pairs for the purpose tanceratios and perpendicu
of graphing. I lar projections permitting the
• seven bridges of konisberg
network
a collection of designated points
connected by paths.
• shape
I exact reconstrUction of a sys
tem of landmarks by a rigid
trusswork. Now, more gener
I ally, coordinates with respect
to any basis for the tangent
space to Kendall's shape space
in the vicinity of a mean form.
the geometric properties of a I
configuration of points that
.. h . I. shape space
are mvanant to c anges m .
. a space in which the shape of
translation, rotation, and I
scale. In morphometries, we a figure is represented by a
represent the shape of an ob I single point. It is of 2p4 di
. mens ions for 2dimensional
ject by a point in a space of ~
shape variables, which are' coordinate data and 3p7 di
~ mens ions for 3dimensional
measurements of a geometric
; coordinate data.
object that are unchanged un
der similarity transforma ; • shape variable
tions. For data that are con : any measure of the geometry of
figurations of landmarks, I a biological form, or the image
there is also a representation I of a form, that does not change
of shapes per se, without any : under similarity transforma
nuisance parameters (posi ~ tions: translations, rotations,
tion, rotation, scale), as single I and changes of geometric scale
points in a space, Kendall's (enlargements or reductions).
shape space, with a geometry Useful shape variables include
given by Procrustes distance. I angles, ratios of distances, and
Other sorts of shapes (e.g., I any of the sets of shape coordi
those of outlines, surfaces, or : nates that arise in geometric
functions) correspond to quite I morphometrics.
different statistical spaces.
155
II shesrlsimilarpolygons
*================
• shear nents" of an allometric analysis
in twodimensional problems, ; of distances to be uncorrelated
shape aspects of any affine : with withingroup size .
transformation can be dia • side
grammed as a pure shear, a map (of a polygon) A line segment
taking a square to a parallelo ; connecting consecutive vertices
gram of unchanged base seg : of a polygon.
ment and height. This is a trans I
formation that leaves one Car : • side of a polygon
tesian coordinate, y, invariant a single segment from the union
and alters the other by a trans I that forms a polygon
lation that is a multiple ofy: for I • sides
instance, what happens when : (of an angle) The two rays, hav
you slide the top of a square ing a common endpoint, that
sideways without altering its form an angle.
vertical position or the length
of the horizontal edges. The I • Sierpinski triangle
score for such a translation, to a type of fractal.
gether with a separate score for : .• significance
change in the horizontal/verti also known as , significance
cal ratio, supplies one orthonor I level, statistiCal significance.
mal basis for the subspace of : The significance level of an hy
uniform shape changes of two pothesis test is the chance that
dimensional data. Without the the test erroneously rejects the
adjective "pure," geometric ; null hypothesis when the null
morphometricians usually use : hypothesis is true.
the word "shear" as an infor I
mal synonym for "affme trans : • similar figures
formation," since any 2D uni two geometric figures are simi
form transformation can be ; lar if their sides are in propor
drawn as one if you wish. In tion and all their angles are the
multivariate morphometries, a . same.
somewhat different use of pure • similar polygons
shear is in a transformation of ; polygons whose corresponding
the "shape principal compo : angles are congruent and whose
I
II
lsimulatiun II
corresponding sides are propor
tional.
6
L
4
r
• similarity transformation
I • simple random sample
I a simple random sample of n
units from a population is a ran
I dom sample drawn by a proce
I dure that is equally likely to
give every collection of n units
from the population; that is, the
I probability that the sample will
consist of any given subset of n
of the N units in the population
I is Iren. Simple random sam
pling is sampling at random
without replacement (without
I replacing the units between
draws). A simple random
sample of size n from a popu
I lation of N > = n units can be
I constructed by assigning a ran
dom number between zero and
one to each unit in the popula
I tion, then taking those units
that were assigned the n larg
est random numbers to be the
sample.
a change of Cartesian coordi
nate system that leaves all ra
tios of distances unchanged.
The term proper or special simi
larity group of similarities is
sometimes used when the trans
formations do not involve re
flection. Similarities are arbi
trary combinations of transla  :
tions, rotations, and changes of I
scale. :
I • Simpson's paradox
• simple events
a single activity in a probability
experiment such as flipping a
com.
• simple polygons
what is true for the parts is not
necessarily true for the whole.
I • ul .
: • Slm ation
·an experiment that has the
; same number of outcomes as a
convex, closed shapes bounded
by line segments joined end to I
end.
given situation but is easier or
more practical to carry out than
the given situation.
Ii sine I slide rule
157
 sine ~ measures of dimension one,
(ofcin acute angle) The ratio of i areas are size measures of di
the. length of the opposite side
to the length of the hypotenuse
in any right triangle containing
the angle.
 singular value decomposi
tion
: mension two, etc.
•
:  skeleton division
~ a long division in which most
i or all of the digits have been
: replaced by asterisks to form a
• ·thm
: cryptan. .
•
· _skew
i two lines are skew if they do not
~ intersect and are noncoplanar.
any mxn matrix X may be de
composed into three matrices
U, D, V (with dimensions
mxm, mxn, and nxn, respec
tively) in the form: X= UDVt, :  skew lines
where the columns of U are ~ noncoplanar lines that don't
. orthogonal, D is a diagonal • intersect
matrix of singular values, and
o
the columns of V are orthogo
nal. The singular value decom •
position of a variancecovari
ance matrix S is written as
S=ELEt, where L is the di •
agonal matrix of eigenvalues
and E the matrix of eigenvec
~  skewed distribution
tors.
 size change factor
size change magnitude
 size measure
in general, some measure of a
form (i. e., an invariant under
the group of isometries) that
scales as a positive power of the
geometric scale of the form.
Interlandmark lengths are size
; a distribution that is not sym
: metrical.
•
.:  slant height
~ the height of each triangular
~ lateral face of a pyramid.
•  slide rule
: a calculating device consisting
~ of two sliding logarithmic
• scales.
Mathematics=================== II
158
slope I sptUe II
• slope I • slopeintercept form
the ratio of the increase in the I the form of a linear equation y
yvalues to the increase in the = fiX + b where m represents
xvalues between any two or I the slope and b represents the
dered pairs. I yintercept.
• slope of a line I • small circle
in a coordinate plane, the the circle formed by the inter
amount of vertical change : sectiCJn of a sphere and a plane
(change in y) for each unit of I that doesn't contain the center
horizontal change (change inx). ; • solid
The slope of a vertical line is the union of the surface and the
undefined. You can calculate the I
region of space enclosed by a 3
slope m of a line (or line seg I D figure; examples: conic solid,
ment) through points with co cylindric solid, rectangular solid
ordinates (xl,yl) and (x2,y2) I
using the formula m =
(Y2  Yl)
( ~  Xl)
• slope triangle
a right triangle used to help
fmd the slope of a line or line
segment through two points,
which are used as the end
points of the hypotenuse. The
length of the triangle's verti
cal leg is the "rise." The
length of the horizontal leg is
the "run." Signs are attached
to each quantity depending on
the direction of travel along
the legs between the points.
• solid geometry
the study of figures in threedi
I mensional space
I • solid of revolution
a solid formed by rotating a
twodimensional figure about a
I line.
I • solidus
the slanted line in a fraction such
as a/b dividing the numerator
I from the denominator.
• space
I in statistics, a collection of ob
: jects or measurements of ob
~ jects, treated as if they were
; points in a plane, a volume, on
the surface of a sphere, or on
any higherdimensional
II ====================Mathematics
II sphere I sttuuJ,srd units
generalisation of these intuitive
Examples are: Eu
clidean spaces, sample spaces,
shape spaces, linear vector
spaces, etc.
• sphere
the locus of pointsin threespace
that are a fIXed distance froma
given point (called the center).
.sphericru oigonometty
the branch of mathematics deal
ing with measurements on the
sphere.
• square
a quadrilateral with 4 equal
sides and 4 right angles.
• square free
159
*================
of deviations between each ele
; ment of the set and the mean
: of the set.
I
: • standard error
the Standard Error of a random
; variable is a measure of how far
: it is likely to be from its ex
pected value; that is, its scatter
I in repeated experiments. The
: SE of a random variable X is
I
: defmed to be SE(X) = [E( (X
I E(X))2 )] V2. That is, the stan
dard error is the squareroot of
: the expected squared difference
I between the random variable
; and its expected value. The SE
: of a random variable is analo
gous to the SD of a list.
an integer is said to be square • standard form
free if it is not divisible by a ; the form of a number expressed
perfect square, n
2
, for n> 1. : as a sum of products involving
• square number powers of ten.
a number of the form n
2
• • standard units
• squareroot law ; a variable (a set of data) is said
the SquareRoot Law says that: to be in standard units if its
the standard error (SE) of the .is z.ero and its standard
sample sum of n random draws dCV1anon IS You transfor.m
with replacement from a box of ; a set of mto standard uruts
tickets with numbers on them : by subtracnng the mean from
IS I each element of the list, and di
• standard deviation
the standard deviation of a set
of numbers is the rms of the set
viding the results by the stan
: dard deviation. A random vari
able is said to be in standard
Mathcmatics======= II
160
. standa1"1lise I stmtifted II
units if its expected value is zero I scale versus interval scale ver
and its standard error is one. I sus ratio scale.
You transform a random vari
able to standard units by sub
tracting its expected value then I
dividing by its standard error.
I • straight angle
an angle whose measure is 180
degrees, forming a line with its
I sides
• standardise
to transform into standard I • straightedge
a tool used to construct straight
units.
lines.
• statistic
: • straightedge, unmarked
a number that can be computed I just how it sounds, an un
from data, involving no un : marked tool used to draw
known parameters. As a func I straight lines
tion of a random sample, a sta
tistic is a random variable. Sta • stratified
tistics are used to estimate pa I this is a feature of an experi
rameters, and to test hypoth mental design whereby a
scheme of observations is re
eses.
I peated entirely using further
• stem and leaf plots sets (strata) of experimental
a method of displaying data I units, with each such further
where the leading digit ( s) are I set distinguished by a level of
the stem and the ending single a categorical variable which is
dio1ts are arranged in ascend I
0 distinct from any categorical
ing order to the side represent I variables used to define the
ing the leaves.
• Stevens' typology
I experimnad design within a
single set (stratum). The data
from the various strata are re
garded as distinct. This situa
tion occurs when attempting
to make inferences based
upon the results of several
similar independent experi
this is widelyobserved scheme :
of distinctions between types of I
measurement scales according :
to the meaningfulness of arith
metic which may be performed I
upon data values. The types are
: nominal scale versus ordinal ments.
II stmriftetlsmnple I stulimt'stcurPe .. 161
• stratified. sample ~ vary enormously with location.
in a stratified sample, subsets ; We might divide the country
of sampling units are selected : into states, then divide each
separately from different strata, ~ state into urban,. suburban, and
rather than from the frame as ; rural areas; then draw random
a whole. : samples separately from each
~ such division.
• stratified. sampling
the act of drawing a stratified
sample.
• stratum
in random sampling, some
times the sample is drawn sepa
rately from different disjoint
subsets of the population. Each
such subset is called a stratum.
(The plural of stratum is
strata.) Samples drawn in such
a way are called stratified
samples. Estimators based on
stratified random samples can
have smaller sampling errors
than estimators computed from
simple random samples of the
same size, if the average vari
ability of the variable of inter
est within strata is smaller than
it is across the entire popula
tion; that is, if stratum mem
bership is associated with the
variable. For example, to deter
mine average home prices in
the India., it would be advanta
geous to stratify on geography,
because average home prices
~ • studentised score
; the observed value or a statis
:" tic, minus the expected value of
~ the statistic, divided by the es
I timated standard error of the
; statistic.
~ • student's t curve
: student's t curve is a family of
~ curves indexed by a parameter
; called the degrees of freedom,
: which can take the values 1,
~ 2, ... Student's t curve is used
; to approximate some prob
: ability histograms. Consider a
~ population of numbers that
; are nearly normally distrib
: uted and have population
~ mean is J.L. Consider drawing
I a random sample of size n with
: replacement from the popula
~ tion, and computing the
~ sample mean M and the
; sample standard deviation S.
: Define the random variable T
~ = (M J.L)/(S/nV2). If the
; sample size n is large, the
: probability histogram ofT can
I
Msth_ti&s======= \I
162
I superimposifitm II
.
be approximated accurately ment of the subset must be
by the normal curve. How ; long to the original set, but
ever, for small and intermedi not every element of the origi
ate values of n, Student's t 1 nal set need' be in a subset
curve with n I degrees of free I (otherwise, a subset would
dom gives a better approxima always be identical to the set
tion. That is, P(a < T < b) is it came from).
approximately the area under I. ••
Student's T curve with n I :  successlve apprOXimatiOn
1 a sequence of approximations,
degrees of freedom, from a to· h cl th
b. Student's t curve can be ; eac one oser to e desired
: value,
used to test hypotheses about
the population mean and con  sufficient condition
struct confidence intervals for 1 a version of a conditional that
the population mean, when tells you when you can use the
the population distribution is term defined, where the term
known to be nearly normally 1 is in the consequent; a condition
distributed. This page con 1 that implies a preset conclusion.
tains a tool that shows _ superimposition
Student's t curve and lets you the transformation of one or
find the area under parts of the 1 more figures to achieve some
curve. I geometric relationship to an
_ subject, experimental other figure. The transforma
subject I tions are usually affine trans
a member of the control group I formations or similarities.
or the treattnent group. : They can be computed by
matching two or three land
 subroutine
1 marks, by leastsquares opti
a previously known algorithm: mization of squared residuals
used in another algorithm at all landmarks, or in other
_ subset ways. Sometimes informally
a subset of a given set is a col ; referred to as a "fit" or "fit
lection of things that belong: ring," e.g., a resistant fit.
to the original set. Every ele I
II supplmuntllry 1.systemRtJ& error
163
• supplementary
two angels are supplementary
of they add up to 18()<>.
~ gram will be symmetrical about
; a vertical line drawn at x=a.
• supplementary angles
2 angles whose measures, when
added together, equal 180 de
grees
• surface
the boundary of a 3D figure
• surface area
(of a solid), the sum of the ar
eas of all the surfaces.
• symmedian
; • symmetric property of
congruence
~ the property of congruent fig
; ures that, if one geometric fig
: ure is congruent to a second fig
~ ure, then the second figure is
~ congruent to the first.
I • symmetry diagonal
: the diagonal that perpendicu
~ larly bisects, the other and is a
~ symmetry line for the kite
reflection of a median of a tri
angle about the corresponding I
angle bisector.
• symmetric distribution
the probability distribution of a
random variable X is symmet I
ric if there is a number a such
that the chance that X> =a + b
is the same as the chance that I
X < =ab for every value ofb. A
list of numbers has a symmet
ric distribution if there is a num
ber a such that the fraction of
numbers in the list that are
greater than or equal to a + b is
the same as the fraction of num
bers in the list that are less than
or equal to ab, for every value
of b. In either case, the histo
gram or the probability his to
I • symmetry line
~ the line of reflection in a reflec
: tionsymmetric figure
I
: • system of equations
~ a set of two or more equations.
~ • systematic error
; an error that affects all the
: measurements similarly. For
~ example, if a ruler is too short,
MnthemRncs============ II
164
=================*
systematic random sample I tltm II
everything measured with it I they were random, if the or
will appear to be longer than I der in which the units appears
it really is (ignoring random in the list is haphazard. Sys
error) . If your watch runs I tematic samples are a special
fast, every time interval you I case of cluster samples.
I • t test
measure with it will appear to
be longer than it really is
(again, ignoring random er I
ror). Systematic errors do not
an hypothesis test based on ap
proximating the probability his
tend to average out.
I togram of the test statistic by
• systematic random sample
a systematic sample starting I
at a random point in the list
ing of units in the of frame,
instead of starting at the first I
unit. Svstematic random sam
Student's t curve. t tests usually
are used to test hypothes.es
about the mean of a population
when the sample size is inter
mediate and the distribution of
the population is known to be
nearly normal.
pIing is better than systematic • t2 statistic
sampling, but typically not as I
a multivariate generalisation of
good as simple random sam I the univariate t
2
statistic. It is
piing. h f h . f th
I t e square 0 t e ratIo 0 e
• systematic sample group mean difference to the
a systematic sample from a I standard error of that differ
frame of units is one drawn by I ence. Used in the 'Ptest.
I • t
2
test
listing the units and selecting
every kth element of the list.
For example, if there are N I
units ifi the frame, and we
want a sample of size N/I0,
we would take every tenth I
unit: the first unit, the eJev
enth unit, the 21st unit, eLc.
Systematic samples are not I
random samples, but they of I
ten behave essentially as if :
a test due to Hotelling for com
paring an observed mean vec
I tor to a parametric mean; or
comparing the difference be
tween two mean vectors to a
I parametric difference (usually
the zero vector). If the obser
vations are independently mul
tivariate normal, then the 'P
II =======MR.themR.tics
\
II :able ojJlalues I tangent space I
test may be used to test null
hypotheses using the Fdistribu
tion. T2 is also closely related
to Mahalanobis D2.
• table of values
a table of two colLUnns, the first
representing values of the inde
pendent variable, the second
representing the values of the
dependent variable.
• tail
165
*=================
I actual outcome as part of the
I tail.
I • tangent
( of an acute angle) The ratio of
I the length of the opposite side
I to the length of the adjacent side
in any right triangle containing
the angle.
I • 1
:. tangent c l r ~ es
I circles that are tangent to the
: same line at the san1e point.
~ They can be internally tangent
I or externally tangent.
an area at the extreme of a
randomisation distribution,
where the degree of extremity
is sufficient to be notable
judged against some nominal I
alpha criterion value.
)
• tail definition policy
this is a defined method for di
viding a discrete distribution
into a tail area and a body area.
the scope for differing policies
arises due to the non
infinitesmal amount of prob
ability measure which may be
associated with the actual
outome value. The conventional
policy, based upon consider
ations of simplicity and of con
servatism in terms of alpha, is
to include the whole of the
I
: • tangent line
~ a line that lies in the plane of a
; circle and that intersects the
: circle at exactly one point (the
~ point of tangency).
~ • tangent segment
; a line segment that lies on a tan
: gent line to a circle, with one
~ endpoint at the point of tan
I gency.
I • tangent space
weight of outcomes equal to the : informally, if S is a curving space
~ and P a point in it, the tangent
MRth_ties================== II
166
========*
tangmtitU veIoeity I tensor II
space to S at P is a linear space I it left the circle along a tangent
T having points with the same I line.
I • tautology
a sentence that is true because
I of its logical structure.
"names" as the points in S and
in which the metric on S "in the
vicinity of P" is very nearly the
ordinary Euclidean metric on T.
One can visualise T as the pro I • tensor
jection of S onto a "tangent an example of a tensor in
plane" "touching" at P just like morphometrics is the represen
a map is a projection of the sur tation of a uniform component
face of the earth onto flat pa I of shape change as a transfor
per. In geometric I mation matrix. The transforma
morphometrics, the most rel tion matrix assigns to each vec
evant tangent space is a linear I tor in a starting (or average)
vector space that is tangent to form a vector in a second form.
Kendall's shape space at a point A rigorous, general definition
corresponding to the shape of ~ of a tensor would be beyond
a reference configuration (usu . the scope of this glossary, but a
ally taken as the mean of a reasonably intUltIve
sample of shapes). If variation I characterisation comes from
in shape is small then Euclid I Misner, Thorne, and Wheeler,
ean distances in the tangent Gravitation (Freeman, 1973): a
space can be used to approxi I tensor is a "geometric machine"
mate Procrustes distances in I that is fed one or more vectors
Kendall's shape space. Since the in an arbitrary Cartesian coor
tangent space is linear, it is pos dinate system and that produces
sible to apply conventional sta I scalar values (ordinary decimal
tistical methods to study varia numbers) that are independent
tion in shape. of that coordinate system. In
• tangential velocity
I morphometrics, these "num"
f
bers" will be ordinary geomet
(0 an object moving in a circle) I
ric entities like lengths, areas,
The speed of the moving object I
in the direction it would take if : or angles: anything that doesn't
I change when the coordinate sys
tem changes. For the represen
I I = = = = = = = M I I ~
167
II terminal side I tetrahedron
*=================
tation of a uniform component  tessellation
as a transformation matrix, the I an arrangement of shapes
"scalars" of the Misner.:rhorne : (called tiles) that completely
Wheeler metaphor arc the LOvers a plane without overlaps
lengths of the resulting vectors I or gaps.
and the angles among them. A
different tensor representing
the same uniform transforma I
tion is the relative metric tensor,
which you probably know as the
ellipse of principal axes and I
principal strains. This tensor
produces the necessary numeri
cal invariants (distances in the
second form as a function of ; . .
.  test statistic
coordmates on the first form) : . . d h th
directly. Other tensors include a use h
to
. test ypo
the metric tensor of a curving ; eses. dYPOb
t
can
un. hi h
, di . constructe y eCl 109 to re
s ace W c expresses stance·. th ull h h' h th
on the surface as a function of Ject e n ypot eSls. w. . e
th
. h' h c. . value of the test statistic IS 10
e parameters 10 w 1C sun ace I .
. d d h some range or collection of
pomts are expresse an t e T . h
cUrPature tensor of the same sur I a test WIt a
c. hi h th· speCified slgmficance level, the
lace W c expresses e way 10 I .
h
· h th C. "C._II y" chance when the null hvpothes1s
w 1C e surlace lauS awa . . .
fi' I I IS true that the test statistic falls
itS tangent p ane at any in the range where the hvpoth
pomt. . uld b . d' be
: eS1S wo e rqecte must
 terminal side at most the specified signifi
the side that the measurement ; cance level. The Z statistic is a
of an angle ends at
 tesselate
the ability of a regIOn to
tessalate
: common test statistic.
I
:  tetrahedron
a polyhedron with four faces.
; The regular tetrahedron is one
: of the Platonic solids.
I
MR.,hmuJhu======= II
168
• tetromino
a foursquare polyomino.
• the five platonic solids
I that elastic energy stretches
and shrinks in the plane of the
original plate can be ne
I glected.) One particular ver
I sion of this problem an infi
nite, uniform plate con
strained only by displacements
I at a set of discrete pointscan
be solved algebraically by a
simple matrix inversion. In
I that form, the technique is a
the five regular polyhedrons:
regular tetrahedron, regular
icosahedron, regular octahe I
dron, regular hexahedron, and
regular dodecahedron.
convenient general approach
to the problem of surface in
I terpolation for computer
graphics and computeraided
design. In morphometrics,
I the same interpolation (ap
I plied once for each Cartesian
• theorem
important mathematical
statements which can be
proven by postulates, defini
tions, and/or previously
proved theorems.
• thinplate spline
coordinate) provides a unique
I solution to the construction of
I D' Arcy Thompsontype defor
mation grids for data in the
form of two landmark con
I figurations.
I • three types of proofs
direct proof in continuum mechanics, a
thinplate spline models the
form taken by a metal plate
that is constrained at some
combination of points and
lines and otherwise free to I
adopt the form that minimises
bending energy. (The extent
of bending is taken as so small I
a proof in which you state pre
I mises, then use valid forms of
reasoning to arrive directly at a
conclusion.
• threedimensional
I having length, width, and thick
ness (i.e., space)
II =======MRthmulriu
II tied rilniu I trlUlitiolUJl mM1
hometrics
• =========1=6=9
 tied ranks
~ what happens when you kick
; the pinball machine too hard. in a non parametric test involv
ing ranked data, if two data have
tied values then they will de
serve to receive the same rank
value. it is generally agreed that
this should be the average of the
ranks which would have been
assigned if the values had been
discernably unequal. Thus, the
ranks assigned to a set of 6 data, I
with ties present might emerge
as sets such as : 1,3,3,3,5,6 or
1,2,3.5,3.5,5,6. The possibility I
of tied ranks leads to elabora
;  Toeplitz matrix
a matrix in which all the ele
~ ments are the same along any
; diagonal that slopes from north
: west to southeast.
I
all a
l2
an
!lIZa]1 a]Z
a
l3
a
lZ
all
al3
a
l2
aJ3 a
l2
all
tions in the otherwisestandard
tasks of computing or tabulat
ing randomisation distributions
where data are replaced by
ranks.
_ tied values
where data are represented by
ranks, tied values lead to tied
ranks. whether or not data are
rep [resnted by ranks, for any
test statistic the occurrence of
tied values will increase the ex
tent to which a randomisation
distribution will be a discrete
distribution rather than a con
tinuous distribution.
 tilt
the measure of an angle as
compared to a horizontal line;
;  torus
a 3D figure formed by rolling
I a rectangle into a cylirtder and
I bending the cylinder until its
bases meet; a "doughnut".
I
:  traee
~ the trace of a matrix is the sum
; of the terms along the princi
: pal diagonal.
I
:  traditional morphometries
~ application of multivariate
I statistical methods to arbi
: trary collections of size or
~ shape variables such as · dis
I tances and angles. "Tradi
~ tional morphometries" differs
: from the geometric
~ morphometries discussed here
MRtm#Cs==================== II
I transformation II
in that even though the dis I formations are used to put
tances or measurements are I variables in standard units . In
defined to record biologically that case, you subtract the
meaningful aspects of the or I mean and divide the results by
ganism, but the geometrical I the SD. This is equivalent to
relationships between these multiplying by the reciprocal
measurements are not taken of the SD and adding the
into account. Traditional I negative of the mean, divided
morphometrics makes no ref by the SD, so it is an affine
erence to Procrustes distance transformation. Affine trans
or any other aspect of formations with positive mul
Kendall's shape space. . tiplicative constants have a
• transcendental number simple effect on the mean,
I median, mode, quartiles, and
a number that is not algebraic.
other percentiles: the new
• transformation value of any of these is the old
transformations turn lists into lone, transformed using ex
other lists, or variables into I acdy the same formula. When
other variables. For example, the multiplicative constant is
to transform a list of tem I negative, the mean, median,
peratures in degrees Celsius I mode, are still transformed by
into the corresponding list of : the same rule, but quartiles
temperatures in degrees Fahr and percentiles are reversed:
enheit, you multiply each ele I the qth quantile of the trans
ment by 9/5, and add 32 to formed distribution is the
each product. This is an ex transformed value of the 1
ample of an affine transforma I qth quantile of the original
tion: multiply by something distribution (ignoring the ef
and add something (y = ax + fect of data spacing). The ef
b is the general affine trans I feet of an affine transforma
formation of X; it's the familtion on the SD, range, and
iar equation of a straight line). IQR, is to make the new value
In a linear transformation, I the old value times the abso
you only multiply by some I lute value of the number you
thing (y = ax). Affine trans multiplied the first list by:
II tmnsformationnotation I
2
what you added does not affect
them.
 transformation notation
t(P), which stands for the trans
formation of P; also Sk where
the transformation S that maps
(x, y) onto (kx, ky) and k is the
magnitude of that transforma
tion
 transitive property of
congruence
of the translation), and have the
; same length (the distance of the
: translation) .
I
:  translation vector
see translation
 transversal
; a line that intersects 2 others
;  transversible
a network in which all arcs can
be traced without going over
lone more than once
I  trapezium
: a quadrilateral in which no sides
I
: are parallel.
I 'd
:  trapeZOl
I a quadrilateral with exactly one
pair of parallel sides. The par
: allel sides are called bases. A
pair of angles that have a base
; as a common side are called a
the property that, if one geo
metric object is congruent to a
second object, which in turn is ,
congruent to a third, then the
first and third objects are con
gruent to each other.
: pair of base angles.
I
 translation
an isometry under which the
vectors (any of which is a trans
lation vector) between each
point and its image are all par
allel (determining the direction
,
:  treatment effect
I the effect of the treatment on
the variable of interest. Estab
: lishing whether the treatment
,
172
=================*
matmmtgroup I trinomud II
has an effect is the point of an I is less than the sum of the
experiment. lengths of the other two sides.
• treatment group I • triangular numbers
the individuals who receive the numbers of dots that can be put
treatment, as opposed to those I into triangular arrangements;
in the control group, who do I equivalently, sums of consecu
not. tive positive integers beginning
• treatment
with 1.
• triangulate
to divide a polygon into tri
angles
the substance or procedure
studied in an experiment or I
observational study. At issue is
whether the treatment has an
• tridecagon
effect on the outcome or vari I
able of interest.
• tree
a tree is a graph with the prop
erty that there is a unique path
from any vertex to any other
vertex traveling along the
edges.
a 13sided polygon
• trigon
I a threesided polygon.
I • trigonometric ratios and
the unit circle unit circle
a circle on the coordinate plane
I with center (0, 0) and radius 1
unit. periodic curve A curve that
repeats in a regular pattern.
period The horizontal distance
between corresponding points
on adjacent cycles of a periodic
• tree diagram .
a concept map in the form of ;
tl1e branches of a tree. You can :
use it to show the relationships ~
among members of a family of ;
curve.
concepts.
• triangle
a polygon with three sides.
• triangle inequality
the property that states that the
length of any side of a triangle
• trigonometry
I the study of the relationships
between the measures of sides
I and angles of triangles.
I • trinomial
I an algebraic expression consist
ing of 3 terms.
II =======MlJtbem.ries
II tromino I twosided hypothesis test
 tromino
a threesquare polyomino.
'. truncated pyramid
part of a pyramid remaining
after truncating the vertex
with a plane parallel to the
base.
 twin primes
two prime numbers that differ
by 2. For example, 11 and 13
are twin primes.
_ twocolumn proof
a form of proof in which each
statement in the argument is
written in the left column, and
the reason for each statement
is written directly across from
it in the right column.
_ twocolumn proof
a form of proof in which each
statement in the argument is
written in the left column, and
the reason for each statement
is written directly across from
it in the right column.
 twodimensional
having both width and length,
but no thickness
 twopoint perspective
a method of perspective draw
ing that uses two vanishing
points.
173
*================
~ _ twopoint shape coordi
nates
a conveniel1t system of shape
coordinates, originally Francis
I Galton's, rediscovered by
Bookstein, consisting (for two
~ dimensional data) of the coor
I dinates of landmarks 3, 4, ...
: after forms are rescaled and
I ..
: repositloned so that landmark
I 1 is fixed at (0,0) and landmark
~ 2 is fixed at (1, 0) in a Cartesian
: coordinate system. Also re
~ ferred to as Bookstein coordi
; nates or Bookstein's shape co
: ordinates.
I
:  twosided hypothesis test
~ c.f. onesided test. An hypoth
; esis test of the null hypothesis
: that the value of a parameter,
~ /L, is equal to a null value, /LO,
I designed to have power against
: the alternative hypothesis that
~ either /L < /LO or /L > /LO (the
I alternative hypothesis contains
~ values on both sides of the null
: value). For example, a signifI
I cance level 5%, twosided z test
~ of the null hypothesis that the
: mean of a population equals
~ zero against the alternative that
; it is greater than zero would
: reject the null hypothesis for
~ values of
1\
174
=================*
(sample mean) I
Izl=
~ true. A 1YPe 2 error occurs if
; the null hypothesis is not re
: jected when it is in fact false.
>
1.96.SE(sample mean)
I
• type 2 landmark
• twoway table
I a mathematical point whose
I claimed homology from case to
case is supported only by geo
metric, not histological, evi
I dence: for instance, the sharp
a representation of suitable data
in a table organised as rows and
columns, such that the rows rep
resent one scheme of alterna
tives covering the whole of the
the data represented, the col I
umns represent a further
scheme of alternatives covering
the whole of the data repre I
sented, and the entries in the I
twoway table are the counts of :
numbers of observations con I
forming to the respective cells
of the twoway classification.
• type 1 landmark
est curvature of a tooth.
• type 3 landmark
a landmark having at least one
I deficient coordinate, for in
stance, either end of a longest
diameter, or the bottom of a
concavity. Type 3 landmarks
characterise more than one re
gion of the form. The multi
I variate machinery of geometric
I morphometries permits them
to be treated as landmark
I points in some analyses, but the
I deficiency they embody must
be kept in mind in the course
of any geometric or biological
a mathematical point whose
claimed homology from case to
case is supported by the stron
gest evidence, such as a local
pattern of juxtaposition of tis I
sue types or a small patch of :
some unusual histology. ~
interpretation.
• unbiased estimator
an estimator, , that has as its
expected value the parametric
I value, q, it is intended to esti
• type land type 2 errors
these refer to hypothesis test
ing. A Type 1 error occurs when
the null hypothesis is rejected
erroneously when it is in fact I
mate: ..
• unbiased
not biased; having zero bias.
II uncontrolled experiment I uniformS;""cmn='P""onetJ=""t = = = = = = " , , 1 ~ 7 ~ 5
_ uncontrolled experiment
an experiment in which there is
no control group; i.e., in which
the method of comparison is not
used: the experimenter decides
who gets the treatment, but the
outcome of the treated group
is not compared with the out
come of a control group that
does not receive treatment.
_ uncorrelated
a set of bivariate data is
uncorrelated if its correlation
coefficient is zero. 1\:vo random
variables are uncorrelated if the
expected value of their product
equals the product of their ex
pected values. If two random
variables are independent, they
are uncorrelated. (The converse
is not true in general.)
_ uncountable
a set is uncountable if it is not
countable.
_undecagon
an elevensided polygon.
~  undefined term
; in a deductive system, terms
: that are assumed, and assigned
~ no properties, "and whose mean
; ing is derived only from the
: postulates or axioms that use
~ them. In our (Euclidean) sys
I tern, the undefined terms are
~ point, line, plane, and space.
:  uniform shape component
~ that part of the difference in
I shape between a set of con
~ figurations that can be mod
: eled by an affine transforma
~ tion. Once a metric is supplied
; for shape space one can ascer
: tain which such transforma
~ tion takes a reference form
; closest to a particular target
: form. For the Procrustes met
~ ric (the geometry of Kendall's
; shape space), that uniform
: transformation is computed
~ by a formula based in
I Procrustes residuals or by an
~ other based in twopoint
: shape coordinates. Together
~ with the partial warps, the uni
; form component defined in
: this way supplies an orthonor
~ mal basis for all of shape space
; in the vicinity of a mean form.
: In this setting, the uniform
~ shape component may also be
MIIthmuJtics========== II
""1,,,,76===========* unilateralsuiftue I unitarydi'Pisor \I
interpreted as the projection
of a shape difference (be
tween two group means, or
between a mean and a particu
lar specimen) into the plane
~ .  union
I the union of two or more sets
is the set of objects contained
I by at least one of the sets. The
I union of the events A and B is
(or hyperplane for data of di denoted "A+B", "A or B",
mension greater than two) and '1\.UB". C.f. intersection.
_ unit analysis
through that mean form and I
all nearby forms related to it
by affine transformations. For
descriptive purposes, the uni I
I the process of using conversion
factors to change from one
measure or rate to another.
form component is
parameterised not by a vector,  unit circle
like the partial warps, but by I a unit circle is a circle with ra
a representation as a tensor, I dius 1.
in terms of sets of shears and
 unit cube
dilations with respect to a I a cube with edge length 1.
fixed, orthogonal set of Car I
tesian axes.
 unilateral surface
a surface with only one
such as a Moebius strip.
 unit fraction
I with one as a numerator and a
natural number as a denomina
"d I
S1 e, tor.
_ unit of attribute
_ unimodal I
a finite sequence is unimodal if :
it first increases and then de ~
the unit chosen depends upon
the attribute being measured
(e.g., for the attribute "length",
creases.
_ unimodular
a square matrix is unimodular
if its determinant is 1.
_ union of two sets a and b
I a tmit might be "meters")"
 unit square
I a unit square is a square of side
length 1.
_ unitary divisor
the set of elements in A, B, or I
a divisor d of c is called unitary
I if gcd(d,c/d) = 1.
both; written AUB
II ===================MJJth_tics
 unity
one
 univariate
having or having to do with a
single variable. Some univariate
techniques and statistics include
the histogram, IQR, mean,
median, percentiles, quantiles,
and SD. C.f. bivariate.
_ universal statement
a conditional that uses the words
'all' or 'everything'
_ universe
in a Venn diagram, everything
that is outside the sets
 upper bound
177
* = = ~ ~ ~ ~ ~ ~ ~
~ _ variable
i a numerical value or a charac
: teristic that can differ from in
~ dividual to individual. Variance,
; population variance The vari
: ance of a list is the square of the
~ standard deviation of the list,
; that is, the average of the
: squares of the deviations of the
~ numbers in the list from their
I mean. The variance of a random
~ variable X, Var(X), is the ex
: pected value of the squared dif
I ference between the variable
; and its expected value: Var(X)
: = E( (X E(X) )2). The variance
~ of a random variable is the
; square of the standard error
: (SE) of the variable.
I
:  vector
any number above which a func
tion value may approach but not
pass.
~ a quantity that has both magni
 valid reasoning ; tude and direction, usually rep
an argument that reaches its : resented by an arrow with a
conclusion through accepted ~ head and tail. The vector's di
forms of reasoning. I rection is indicated from the tail
_ vanishing line : to the head. The vector's mag
the horiwn; in a drawing it is ~ nitude is the length of the line
at the height of viewer's eye ~ segment.
_ vanishing point I  velocity
a point toward which lines in a ~ the rate of change of position.
perspective drawing converge if : _ Venn diagram
they represent parallel lines that ~ a pictorial way of showing the
recede from the foreground to I relations among sets or
the background. :
178
events. The universal set or I • vertex arrangement
outcome space is usually I a notation that uses positive
drawn as a rectangle; sets are integers and other symbols to
regions within the rectangle. I describe the arrangement of
The overlap of the regions I regular polygons about verti
corresponds to the intersec ces of a semiregular tessella
tion of the sets. If the regions tion. Also called the numeri
do not overlap, the sets are I cal name.
disjoint. The part of the rect I • vertex of a conic solid
angle included in one or more the point that marks the thin
of the regions corresponds to I nest part of a conic solid
the union of the sets. This
page contains a tool that illus • vertex of a polygon
trates Venn diagrams; the tool ~ an endpoint of a segment in a
represents the probability of ; polygon
an event by the area of the • vertex of an angle
event. the common endpoint of the
• vertex
I two rays
I • vertical angles
nonadjacent angles formed by
the intersection of two lines .
• vertical line
I a line that goes straight up
and down, and whose slope is
defined as infinite or unde
I fined .
(of a polygon) The point of in I • vertices
tersection of three or more I plural form of vertex; the point
edges. of intersection of the rays of an
I angle, "corner" point of any
geometric figure bounded by
lines, planes, or lines and
• vertex angles
(of a kite) The angles between I
the pairs of congruent sides.
planes.
II =======M#them4ties
11l'iew I test
179
• view • whole number
a drawing of a side of an object ; a natural number.
• vigesimal
; • Wilcoxon test
related to intervals of 20. : named after the statistician F,
• vinculum Wilcoxon . This test applies to
the horizontal bar in a fraction I an experimental design involv
separating the numerator from ing repeated measure ob
the denominator. : servanons on a common set of
I experimental units, which need
• volume
the amount of space a 3D ob
ject can hold.
• vulgar fraction
a common fraction.
be only ordinalscale. the pur
: pose is to measure shift in scale
I location between the two levels
of the repeated measure distinc
: tion. the test statistic is derived
from the set of differences be
• weak inequality ; tween the two levels of the re
an inequality that permits the : peated measure distinction one
equality case. For example, a is difference score for each obser
less than or equal to b. ; vational unit. the procedure is
• weight matrix : somewhat variable between
w matrix :fhe matrix of partial authors, although the variants
warp scores, together with the I each correspond to valid well
uniform component, for a : sized exact tests. Wilcoxon's
sample of shapes. The weight original procedure commences
is as a I by discarding entirely the obser
non of the ProcrustesresIdual vations from any experimental
shape coordinates; like them, : units for which the data values
they are a set of shape coordi are equal at each level of the
nates for which the sum of ; repeated measure comparison.
squared differences is the : thus or otherwise, the next step
squared Procrustes distance be is ranking the differences, pro
tween any two specimens. ; viding a rank for each retained
wff : experimental unit; the ranks are
• according to the absolute values
a wellformed formula.
Mtldlmufriu======;;;;;;;;;;; II
180
wintling number I xpentoltlitllJ II
proceeds to fmd additional fac
summed separately into two or ; tors to fit to the residuals, and
three categories: negative dif : so on until the data is ad
ferences; zero differences (if ~ equately fit.
any); positive differences. the ; • x
of the differences. The ranks are I
test statistic is the smaller of the
I roman numeral for 10.
outer categories, plus an adjust
ment for the middle (zerodif I • xaxis
ference) category. the horiwntal axis in the plane.
• winding number • xintercept
the number of times a closed the point at which a line crosses
curve in the plane passes around I the xaxis.
a given point in the counter I • XOR, exclusive disjunc
clockwise direction.
tion
• witch of agnesi
a curve whose equation IS
x
l
y=4a
2
(2ay).
• work
XOR is an operation on two
I logical propositions. If p and q
are two propositions, (p XOR
q) is a proposition that is true
I if either p is true or if q is true,
but not both. (pXOR q) is logi
cally equivalent to «p OR q)
I AND NOT (pANDq».
the result of force applied over
some distance, given by the for I
mula w = fd, where w = work,
f = force, and d = distance.
• Wright factor analysis
I • xpentomino
a version of factor analysis, due I
to Sewall Wright, in which a I
path model is used to describe
the relation between the mea, ~
sured variables and the factors I
of interest. It is usually explor
atory, in that one fits a simple
one factor model iteratively to I
maximally explain the correla
tions among variables, and then
a pentomino in the shape of the
letter X.
181
* = = ~ ~ = = = = = = = = ~
• yard ~ nal)jSE(original).
a measure of length equal to 3 ~ • zero
feet. ; 0
• yaxis
the vertical axis in the plane.
• year
a measure of time equal to the
period of one revolution of the
earth about the sun. Approxi
mately equal to 365 days.
• yintercept
the point at which a line crosses
the yaxis.
• z statistic
a Z statistic is a test statistic whose
distribution under the null hy
pothesis has expected value zero
and can be approximated well by
the normal curve. Usually, Z sta
tistics are constructed by
standardising some other statis
tic. The Z statistic is related to
the original statistic by Z =
(original expected value of origi
; • zero angle
: an angle whose measure is O.
~ In a zero angle, both the initial
; and terminal sides are the same.
; • zero divisors
: nonzero elements of a ring
~ whose product is O.
~ • zero element
; the element 0 is a zero element
: ofagroupifa+O=aandO+a=a
I
: for all elements a.
I di . I
.• zero menSlona
~ having no dimension; a point
I • zintercept
~ the point at which a line crosses
: the zaxis.
I
: • zone
~ the portion of a sphere between
; two parallel planes.
; • zscore
: the observed value of the Z sta
I risric.
I • ztest
; an hypothesis test based on ap
: proximating the probability his
~ togram of the Z statistic under
; the null hypothesis by the nor
: mal curve.
I
_cP!_ COMPREHENIIVE vvvvOfvvvv
[)ICTI()~Al?"
MATHEMATICS
Chief Editor & Compiler:
~
Roger Thompson
~
ABHISHEK
All rights reserved. No part of this book may be reproduced in any form, electronically or otherwise, in print, photoprint, micro film or by any other means without written permission from the publisher.
ISBN
Copyright
9788182473416
Publisher
Revised Edition
2010
Published by ABHISHEK PUBLICATIONS, S.C.O. 5759, Sector 17C, CHANDIGARH1600 17 (India) Ph.2707562,FaxOI722704668 Email: abhpub@yahoo.com
or a general reader with science curiosity. number theory. Featuring more than 1500 terms. . scientists. and other areaseach with definition along with pictorial representations of many terms. this comprehensive resource defines the current language of pure and applied Mathematics and gives you a better understanding of the ideas and concepts you need to know. and students will look this dictionary for defmitive coverage of all branches of mathematics. This dictionary will be a handy reference for the mathematician or scientific reader and the wider public interested in who has said what on mathematics. engineers. It has. both pure and applied. topology. This dictionary is authoritative. comprehensive comprising latest terms and is carefully reviewed to ensure its accuracy. fmite mathematics. a student. Whether you're a professional. clarity. brought together in one easily accessible form the bestexpressed thoughts that are especially illuminating and pertinent to the discipline of Mathematics. operator theory. the book defmes terms and expressions in algebra. logic.Preface Mathematicians. and completeness. for the first time. complex numbers. a writer. This Dictionary of Mathematics puts a wealth of essential information at your fingertips.
clarifies and provides power. The subject has a reputation for being disagreeably difficult.· The overall aim of the dictionary is to provide an accessible description of what one judges to be the core material for damn good dictionary. . I have tried to alter that perception by showing that key ideas can be presented simply judiciously but not overwhelmingly deployed.
11 2by2 table I abeliangroup
5
.2by2 table 1. this is a twoway table where the numbers of levels of the ro·.v and columnclassifications are each 2. If the row and column classifications each divide the observational units into subsets, then it is likely that it will be useful to analyse the data using the Fisher Test. • 3D figure a set of points in space; examples: box, cone, cylinder, parallelpiped, prism, pyramid, regular pyramid, right cone, right cylinder, right prism, sphere.
. ; .
I
I
T
0.
A~
45'
I
~ .8
; 'prime'; designates an image : corresponding to the preimage ~ using the same variable.
I.
I
.• aRb I a is an element in b
I
abacus ~ a Japanese counting device and : calculator.
~ • abelian group
a group in which the binary
.'1...
f
°',;'"~.if..... rIO;'.if. ...... (~, '=) .. .._'1 . •.
w;;g,
Wi'Z"
• 454590 triangle an isoscoles right triangle
!!!!!6="
========!!!!!!!!*
~
sbseisss IIIJlinetnlnsj'tmnllritm
II
is commutative, that IS,. ab= ba for all elements a abd b m the group. • abscissa the xcoordinate of a point in a 2dimensional coordinate systern . • absolute value the positive value for a real n~ber, disregarding the sign. Wntten 1x I. For example, 131 =3, 141 =4, and 101 =0. • abundant number a positive integer that is smaller than the sum of its proper divisors. • acceleration ~e rate of change of velocity With respect to time. • "'.........hl I els f 1'~ e eVi 0 accuracy ~e p~eClSlon determined by the sItuation or the given numbers stude?ts should help develo~ what IS acceptable according to the situation.
?peratio~
• additive identity property ; the sum .pf any number and : zero is the original number· ~ zero is the identity element of ~ addition. ; • adjacent angles . . g ht and nonzero ; two nonstraJ. : angles that have a common side ~ in the interior of the angle ; f?rmed by the nons:ommon : SIdes
~
~ ... ~" Be 0 DE
~
lnIeriorangl....nlhe ....... side oflhe InirIlIversaI A
IntedonI on the ......
1IIC3o_18D
::~
; . . ; • ad)a~t interior angle : the mtenor angle that forms a ~ linear pair with a given exterior I angle of a triangle. : . ." adjacent SIde ~ (of an angle.in a triangle~, one : of the two SIdes of the mangle ~ that form the sides of the angle.
I·
• accuracy ; • affine superimposition the closeness of a measurement ; a su~rimposition for which the or estimate to its true value. : assocIated transformations are I all affine. • acute angle : an angle whose measure is ~ • affine transforma,tion greater than 0 but less than 90 ; a transformation for which degrees. : parallel lines remain parallel.·
I
~~~~~~~~~~==~~~M
llilleeimlie elJllllhon lilleeimlie
numb:.. =========~7
was applied ; based shape .
I
Mfine transformations of the plane take squares into parallelograms and take circles into ellipses of the same shape. Mfine transformations of a 3dimensional space take cubes into parallelopipeds (sheared bricks) and spheres into ellipsoids all of the same shape. Similar results are produced in higher dimensional spaces . Equivalent to "uniform transformation" . As far as form is concerned' (that is, ignoring translation and rotation), any affine transformation can be diagrammed as a pure strain taking a square to a rectangle on the same axes. In studies of shape, where scale is ignored as well, the picture is the same but now the sum of the squares of the axes is unchanging. Still ignoring scale (that is, as far as shape is concerned), any affine transformation can be also diagrammed as a pure shear tak ing a square into a parallelogram of unchanged base segment and height. This diagram of shear came into morphometrics via an application to principal components analysis somewhat before it
~
to
landmark
'g
I
'. Y 'II
:or
!r<7H:~,
_:l____ ~'
I
L ' .1, "8;,
~.::~~ .
. '
: .t,. ....
I
x,.
x"
~  algebraic equation
I
an equation of the form
~ f( x) = 0 where f is a polyno
: mial.
I
tPq
= Wq • 1l.y • tlq = q. eq = (ra  i) . ax + q.m = !6f. st n
where
I
Wq
I
q
tlq
I
and
I I
6=
(~)!
: _ algebraic number ~ a number that is the root of ; an algebraic polynomial. For : example, (sqrt 2) is an alge~ braic number because it is a ~ solution of the equation x2 =2.
.MiJt'nes~====== II.
clear . · 2 '1 . tdgorithm I RltemateinterWrangles {~J.I y ate a falsepositive error : affirming a nonnull pattern by chance. an algorithm expressed in a I this present glossary.. ~. .!!!!8~~~~~~~~~~. a statistical test will gener. 11' to. I tistics which are the concern of 2.. {~. 11.alpha also known as size or typel er. {i. in advance of undertaking the test.'. Ti. ~}.21J. complete and unambiguous... 11..}' 11} esis) is no greater than this _ algorithm nominal alpha criterion level. I . 0.05). Tz' T. 1. 1 ing statistical signifilCance if the · 1 AlqebCnct.. l.I u"'. The outcome under the null hypoth. rz}' 11 . 'I'. I actual alpha (probability of the 1. a cryptarithm in which the let_ allometry I ters. 2I1'.alternate interior angles if two parallel lines are cut by a ror. bivariate relation from the .'1· .}. 11' •. I if two parallel lines are cut by a where c is a constant ~d x and I transversal.~. 11 . {i. ~ terion level (often 0. U1C(.alphametic computer .. alternate exterior y are size measures in units of : angles are outside the parallel I lines and on opposite sides of the same dimension. a formal statement. to set a nominal alpha cri.2I1'. ~ This reasoning is applicable for all types of statistical testing. IT' + + IT}' 11. according to some null hypothesis.. {o.I . of I how a certain process needs to : including rerandomisation stabe undertaken. programming language for a ."'. Conventional method. IT'~' Ti}. i. II .1. .19ttbrUcl'·Z11' Co. form related words or It describes any deviation of the meaningful phrases.alternate exterior angles simple functional form yIx = c.""cl... 0.'I'.'1· •• {~. outcome is classified as showU .1"""<1> 211'. I the transversal. This is the probability that.I .'z ology for statistical testing is. . which represent distinct any change of shape with size.cl. u""0. I digits.
nor that the alternative hypothesis is true. in evaluating whether a new cancer remedy works. while the alternative hypothesis would be that the remedy does work. changing I • amicable numbers . the null hypothesis typically would be that the remedy does not work. the other. : ~ ~ or the line containing the opposite side. For example.9 *================= transversal. two numbers are said to be : amicable if each is equal to the ~ sum of the proper divisors of . a null hypothesis (typically that there is no effect) is compared with an alternative hypothesis (typically that there is an effect. When the data are sufficiently improbable under the assumption that the null hypothesis is true.tween the planes : of the bases I • ~ ~ • altitude of a trapezoid I ~ : I : I . the perpendicular segment from a vertex to the line containing the opposite side of a triangle : • ambiguous ~ not stable. • altitude of a conic solid the length of a segment whose endpoints are the vertex and a point on the plane of the base that is perpendicular to the plane of the base altitude of a cylindric solid the distance be. nor that the null hypothesis is false. the null hypothesis is rejected in favour of the alternative hypothesis. alternate interior angles are inside the parallel lines and on opposite sides of the transversal.I logic of a situation or a process. pendicular to the opposite side . • alternative hypothesis in hypothesis testing. . • analyse • altitude : to break down into parts and (of a triangle) A line segment ~ explain or demonstrate the drawn from a vertu that is per. : I the distance between the bases of a trapeziod • altitude of a triangle. or the length of the altitude segment. : ~ ~ . or that there is an effect of a particular sign). ~ . (This does not imply that the data are probable under the assumption that the alternative hypothesis is true.
. (of a polygon) an angle having its vertex at one of the polygon's vertices. with respect to time.10 angle I anisotropy I II • angle 1. In two dimensions. I • I I I I I • I I • angle side one of the two rays forming an angle • angular velocity I (of an object moving around a circle) the rate of change.: ject and a ftxed point. Also. and that divides the angle into two congruent angles. • angle bisector a ray that has the vertex of the angle as its endpoint. and the viewer's line I • anisotropy of sight to the object as the anisotropy is a descriptor of one other side. two noncollinear rays (the sides of the angle) having a common endpoint (the angle's vertex). angle ruler a hinged ruler with a protractor attached for reading the measure of an angle in degrees. the measure of the angle. and sides that go through the point and its image. one side I horizontal. and having two of the polygon's sides as its sides.\. tex at the viewer's eye. 2.formed by an affine transfor . with its vertex at the center of the rotation. one side horizontal. of the measure • angle of depression I of the central angle that inter(of an object lower than the I cepts the arc between the obviewer) The angle with its ver.ooe •• _ I I tex at the viewer's eye. LI!OH'''' . and the viewer's line of sight to the object as the other side. this is • angle of elevation the ratio of the axes of the el(of an object highet than a I lipse into which a circle is transviewer) The angle with its ver. I aspect of an affine transformation. angle of rotation the angle between a point and its image under a rotation.
measured in units of length. ~ and connecting the vertices so I that the lateral faces are tri~ angles. .: rotating one base of a prism tric circles of unequal radius. In general. the polygon's circumscribed : circle to a side of the polygon ~ that is also perpendicular to .riu======= II .apothem : (of a regular polygon) a line ~ segment between the center of . From Latin for "before. Also. ~ . The specifications are generally arranged to concur with those of ISO. This body publishes specifications for a number of standard programming languages. _ ante the upfront cost of a bet: the money you must pay to play the game. same endpoints. I _ ANSI acronym for the American Nationa! Standards Institute. represented by p. problem I •• tlon. An angle interI cepts an arc if the sides of the ~ angle ~tersect the circle at the : ~ndpomts of the arc. the length of : that line segment."." : .antlpnsm I a polyhedron resulting from . ~ .: esis. Th~ arc is ~ mcluded by the chord wlth the .. that side. the 'if' part of a conditional' l~n~ in one direc~on to exten. aka hypoth: s~on m a perpendlCular direc. it is the ~ _ antecedent maximum ratio of extension of .arc ~ (of a circle) Two points on the .IllInnulus I lire length 11 mation. MII. : .annulus the region between two concen. circle (the endpoints of the arc) : and the points of the circle be~ tween them.arc length : the portion of the circumference ~ of the circle described by an arc... given.
which is a special case of association in which large values of one variable tend to occur with large values of the other." "'" I B '. . asymptote a straight line always approaching but never intersecting a curve. // . u ..dAC . the two variables are associated. · '"' I ...ID< /""~o . in vertical "slices") than it is overall. if the scatter in the values of the variable plotted on the vertical axis is smaller in narrow ranges of the variable plotted on the horizontal axis (i.e. .~.00 I I . ' I "'. The correlation coefficient is a measure of lillear I • I I • I • automorphism an isomorphism from a set onto itself. I I I • assume to accept as true without facts or proof..12 • arc measure the measure of the central angle that intercepts an arc.. M J J .0.7•. average a sometimes vague term. In a scatterplot of the two variables..Sl. and small values of one tend to occur with small values of the other (positive association). and vice versa (negative association)./. or in which large values of one tend to occur with small values of the other. . \ / _. the c formula (a + b) + c = a + (b + c).. ..4. I I I I ... association. associative property property about grouping of numbers. I • • area the amount of space taken up in a plane by a figure • arithmetic mean the arithmetic mean of n numbers is the sum of the numbers divided by n.. automatic drawer a computer program that lets you build constructions • association two variables are associated if some of the variability of one can be accounted for by the other. / ". measured in degrees. of addition. It usually denotes the arithmetic I • II = = = = = = = . " . / / ...dA( .
All other mathemati~ cal facts about probability can I be derived from these three axi~ oms.I 2. any parallel pair 100%. for ~ 3. : erence when drawing an altiThe second axiom says that the I tude or other feature. For example. the geometric mean. the line segment lind ) I Th I east z~ro. but it can also denote the median. there are three axioms of prob. and weighted means. If the faces are its edge or doesn't land at all is ~ all rectangular. (ofa cY er. For example. con. among other things. the· mode. 4. This is a : consequence of the second and • axioms of probability ~ third axioms. the same time (if they are dis. ~at the coin lands heads. If I bases two events cannot both occur at : . can be considered the bases.I • axis ability: Chances are always at . I. e c~ance that : connecting the centers of the something happen! IS 100%. The third axiom says. it is true that : the chance that an event does ~ not occur is (100% the chance . chance that the coin either lands . that the event occurs). must be at least zero.~ • base joint or mutually exclusive). is the sum of the chance that the : coin lands heads and the chance ~ that the coin lands tails. the congruent parallel polyheads or lands tails or lands on : gons of a prism. a side of a polygon or face of sider an experiment that con.: a solid used for reference when sists of tossing a coin once. because I both cannot occur in the same : coin toss. the . that the chance that the coin eimean. a face of a solid used for refmstance. ~ . the side of an isoscoles trichance that either one occurs is : angle whose endpoints are the the sum of the chances that ~ vertices of the base angles each occurs. The ~ drawing an altitude or other first axiom says that the chance I feature.13 ther lands heads or lands tails .
. aXIS 0 the mean landmark . . 1 ocanons . 2. (of a trapezoid) A pair of I flat metal plate. n genera. ~ there i~ an alternate formula • Bayes' rule Bayes' rule expresses the con. s ape c anges IS e).I nme._ d : one of these outofplane madcs that are asslgne to llAe I " h h )). th (·deal '. infinitely thin. x is called : xP(Ac». the nonrectangular ~ P(AIB) = P(B IA) xP(A)/( pair forms the bases. din f connecting the pair of land.o that the landmarks ..: that remains meaningful in ditional probability of the event ~ morphometries: bending enA given the event B in terms of i ergy is proportional to the inthe conditional probability of : tegral of the summed squared the event B given the event A: ~ second derivatives of the "ver . e n g energy 0 . dina coordinates lor an ar m a I e · . . Whil· h· bending m shape and pass near the centroid i. measur m o a mean s ape . i P(B IA) xP(A) + P(B lAc) 5. appropriate units (g cm2 sec2). d 1::. the base and y is the exponent.~ that the differences in coordi: nates ~ these same ¥dmarks ezoid as a common side. IS r. in the expression xY. IS a r quannty. I dm ks· : plate perpendicular to Itself. eeal p YSI~ eneder~ f th t h . (0 0) d (1 0)· th . • b Ii I in ano er picture are taken as ase ne : . ed be d th tal I baselines work better if they are i qwr to n e me pate closely aligned with the long : s. and suppose angles with a base of the trap. (of an isosceles triangle) The I Imagine a configuration of two angles opposite the two : landmarks that has been printed ~ on an infinite.14 ~~~~~~~~. th lin artesIan te at a I theaselinelS e e : one Th be coor b pane. 3. me·I ized) energy that would be reI constructIon. ~ • bending energy • base angles ~ bending energy is a metaphor 1. (of an isosceles triangle) The i borrowed for use in two angles opposite the two : morphometrics from the mecongruent sides. congruent sides. Otherwise.were . A: I f tho for a system of twopoint shape I vernc 'tf'~ a~ements o. f I lifted or lowered appropnately. anI . ~ chanics of thin metal plates.
pattern of deviation from a null (usually wntten Lk·l) deter.II hentlinamewmlltriJ&lhettl . single inversely pr?pomonal to scale. pl~ any two perpendicular coordi.'p'. That is. evolutionary distance) between .g:. The bending energy of a ~ of ~e bending energy of Its shape change is the sum of the : honzontal:co~~nent. taxonOIll1C or ~ possible outcomes.error. probability that a statistical t~t portional to that mtegral of : will generate a falsenegative those summed squared.: a coin toss. of an affme transformation is ~ • Bernouilli process zero since it corresponds to a : this is the simplest probability tilting of the plate with?ut any ~ model a single trial between bending. then I level the corresponding level for bending energy is htLklh. This is the the formula whose ~alue IS pro. The distribution . representmg the Sl~ch quantities should not probability attributed to one interpreted as measures o~ dis. ta • bending energy ~atrix : is the complement to power : the formula for bending ~nergy ~ beta = (Ipower).~ the bending energy of Its v~~ nates in which the metaphor is : cal ycomponent. ~ parameter. The value obtamed for I two possible outcomes such as the bending energy correspond. Conlandmarks of the reference I ventional methodology for staform.' 2 be I also known as type. modbending energies that apply to I eled as ~ vemcal pl~te. modeled Slffilevaluated. ing to a given ~splacement is ~ depends upon a .: pattern in circumstances where mined by the coordinates of the ~ the defmed pattern exists. I M s t h _ t i c s = = = = = = II .: defmed outcome out of the two similarity (e. second ~ error: failing to assert a defmed derivatives is. tion is the sum xtLklX +y'Lk:ly tilt. The bending energy I !arly as a "vertical" plate. a quadratic form . ~ • beta two forms. In : BETA will depend upon the morphometrics. the bending ~ nominal alpha criterion level ?e . .~~~~~~~~~15 tical" displacement the extent to ~ energy of a general transformawhich it varies from a uniform.. if h is a vector : tistical testing is to set in addescribing the heights of a plate ~ vance a nominal alpha criterion above a set of landmarks.
...... ~..~I ....~~:~~_"+I""''' _.1'IWJ'>(.l p.... .:c!W~").... IAJ pI"D ".•• .~ I ...os.."~~.. r.J example._ _ .. Applicability of this rea. It is question.  (2AIOJ1J:"'=.... .th.... 1d"'{~U rd'fr... : I • I .I0(h3 ~~.....1 .l { lAl .ynumber II and upon further considerations J swer that differs from the truth.....) .:.(i)'~~ (lao::: . I ...~s!:i'" ttel!'" a:u. For ally in the relative power of dif. ~~~t:¥ :=:=::. ...2 • bias a measurement procedure or estimator is said to be biased if. including the strength of the J The bias is the average (expattern in the data and the pected) difference between the sample size.. but it is not necesfrom the concerns of this Glos.: measurement to be larger than able whether the concept of J your true weight (this would be BETA error is properly appli.r~J••• .16 bias I biMr..~ ~ t:'41 6:1 rt.....Q:d.fUIJ~. ..>'1t'.(. uses the words if and only if..\I) 11~ I 1 I • bimodal having two modes. on the average.a 1loCt.t[!p. l '..Jt'" I\~_'_"~~&IJx)..I .. _ . •• ...~~_~~L_ t!ffllW~ .. (I·/b"SV..t""...._Wool r.. .I • biconditional soning is also closely bound up I Q conditional and its converse with the choice of test statistic. "'... .' •• m .'...4~1/ It<W1 (J...'~~5~ =.=r... .."IQt3"'I""~""''' .). .. bilateral symmetry reflectional symmetry with only one line of symmetry..: a positive bias).' 32 . if you get on the scale ferent tests rather than in an with clothes on..p}.l#t.add ..'. ""'' 'O''.....". WltS) J·. • tIII OIl ..i." . The design of cable without considering the an experiment or of a survey can concept of sampling from a J also lead to bias. uacD rJwJ·Mi.. . r.. t14tiIJ .J measurement and the truth. ih:::::" Id~'!!. that biases the absolute value... ~.. ~ where the converse is also true.)a~ IiaJtJalnn 1umi~ _?:ial. _&1Jtt /~.~ I I I I toueo...a. Interest is gener.=~ ::=::. • . binary number a number written to base 2.."" ~~c...... Bias can be population... ~ i=:=::'...~_::/ :=5... .  . ::::1Ia. r :.. .m:.__ I . sary... 2 ... 0. :. a.. which is separate deliberate.sarily so.""IJ.""" I... 4. it gives an an J I  :. ~ r. .. I • I ~.~...J ::::==. • .~.:r .).. • bijection a onetoone onto function.
A value for the parameter 'p' is hypothesised (null hypothesis) and the difference of the actual value from this is assessed as a value of alpha." Under these assumptions. : ~ . traction are binary operations. . The expected value ~f a r~do~ v~iabl~ wi~ the Bmomial dIstnbutIon IS nxp. operation I biseet • binary operation ~ a binary operation is an opera. an expression that IS the sum of two terms. all of which have the same probability p of resulting in "success. . : • biplot a single diagram that represents two separate scatterplots on the same pair of axes. and the. : I . •. the probability of k successes (and n _ failures) is k nC pk(lp)nk.II hiM". ~ • biquadratic equation . : bmolD1~. bl~e~t • . this IS a stansnc~ test refemng to a repeated bmary process such as would be expected to generate outcomes with a binomial distribution. When S is a centered data ~ ~atrix. One scatter is of some pair of columns of ~ the matrix U of the singular I value decomposition of a ma: trix S.I tion that involves two operands. . • binomial distribution a random variable has a binomial distrib. • binomial test .stanruu:d error of a random vanable With the Binomial distribution is (nxpx(l p» V2. . a polynomial equation of the : 4th degree. to diVide mto two congruent I parts. where nC is the k k number of combinations of n objects taken k at a time: nC = k n!j(k!(nk)!). the effect is to plot prin: Clpal component loadings and ~ scores on the same diagram. . . I 17 probability histogram of the binomial distribution. ~. I ~ . . • binomial coefficient the coefficients of x in the expansion of (x+ 1)n.:!:tion (with parameters n and p) if it is the number of "successes" in a fixed number n of independent random trials. and the other scatter is ~ of the matching pair of columns I of V. For example. This page shows the : ~ ~ .. addition and sub.
For example. . .th g of experimental conditions). onte. The positive motive for ~ bootstrap resampling is the gen 1I=======MAthemsties . and regression ~ nmes d ~ a atum ~ay inglre make sense for bivariate data . . d··d ua1"· a f at h erson .h ". di ·dual"· ) th I randomisation scheme.. resamp su set. ume 0 f gas un er a certam set. In order to have a blind experiment with fixed arbitrarily independently : of the parameter values of the ~ experimental design and may . WI o~ersamp th erebels nO · · constramt upon e num r 0 f etc. it is usually any plane. The distm. necessary to administer a plasional figure containing the the cebo to the control group." These measurements I sampled. asf leatures w c d songws . e pressure an tem. the subjects do not know whether they are in the treatment group or the control group. M e l0 test. bivari. the bootstrap procedure hel ts 0 f fathers and sons (an I h Co hi h di· .I placements as the data are dividual.I subsets of the data on the basis ate data are data where we have of random sampling with retwo measurements of each "in.18 bisecWrpflJ~t I bootstmp II • bisector pf a segment ~ human subjects. midpoint of the segment and no I • bootstrap other points on that segment I this is a form of randomisation • bit I test which is one of the alternaa binary digit.. tha d be p coefficient. Such resampling promight be the heights and vides that each datum is equally in the weights of a group of people I represented ". how(~ghID VI IS a person. d . perature of a fIXed volume of I ". the resampling subsets may be • blind experiment in a blind experiment. e: ever. di . even exceed the total number of : data.ar . the SIZe 0 f but not uruvariate ata. strap procedure are concerned lin th . Scatterp1 the corre1 ots. point or twodimen.. resente lingID genberaong a ~ e . tives to exhaustive re• bivariate randomisation. anon·. III IVI IS h f ... Th~ bootstrap having or having to do with two I scheme involves generating variables.dual"· th 1· gwshing features of the boot(an ID VId gas IS ~ vo .) th d I It rom t e proce ure 0 a pair..
SD(box) . : where for a box that contains ~ only zeros and ones. A negative aspect of the bootstrap is that the form of the resampling distribution with prolonged resampling converges to a form which depends not only upon the data and the test statistic. I ~ the sample to estimate the SE . : . For I sample percentages. with re~ placement. tickets "at random" from a box use the SE for sampling from : with replacement.II bootstrapestimateofstandarderror I: eral relative ease of devising an appropriate resampling algorithm when the experimental design is novel or complex. the SE of the sample percent: age consists of estimating ~ SD(box) by ((fraction of ones I in sample) x (fraction of zeros : in sample»'h. When the ~ sample size is large. model 19 . this apI proximation is likely to be ~ good. the SD of the . but also upon the bootstrap resampling subset size thus the resampling distribution should not be expected to converge to the gold standard form of the exact test as is the for MonteCarlo case resampling. for the purposes of es: an analogy between an experitimating uncenainties.box ~ a surface made up of rect~ angles. sample is the bootstrap estimate : of the SD of the box from ~ which the sample is drawn. . that the ~ ment and drawing numbered sample is the population.( (fraction of ones in : box) x (fraction of zeros in box) ~ )Y2." which connotes getting out of a hole without anything to stand on. this takes ~ a particularly simple form: the : SE of the sample percentage of I n draws from a box. epiped Msthemsties=========~ II . The bootstrap estimate of . of sampling from the popula: tion. a rectangular parallel. For sampling from a box ~ of numbers. then . The idea of the bootstrap is to ~ • box model assume.bootstrap estimate of standard error the name for this idea comes from the idiom "to pull oneself up by one's bootstraps. is SD(box)/nY2. An effective necessity for the bootstrap procedure is a source of random codes or an effective pseudorandom generator. For example.
the number 0 f sub' Jects : ranc an . time. treatment are a random sample ~ a dot represents each of the high of size n/2 from the set of n I and low values of the data. . with than the median time. box with two tickets in it: one . R or a placebo to a group of n in.~ a horizontal line called a whister in less than median time. to simplify things. . anOclpate the I ' the next f nl2 . and half 10 more ~ the upper half (quartile 3).oun . . and that n is' even). 'd w h0 recelve the remedy and I exploratIon' a randOlrusanon " ..~ the median of the lower half got better in less than. rannOffilSatIOn 're atIveThis ale to draws Wlth d . If the remedy is ineffec. present r~ OffilSanon.: dividuals. a representation of data above covered m exactly the median: a numbered scale where the time.~ N one recovers from the cold : eventually. m the context 0 f a ~ randomisation test such selec=~~~~~~=* II =======MsthtmUJms .re. suppose we are trymg to evalu. half the individuals . half of whom got bet.: (quartile 1) and the median of dian time. .20 . Consider the me.~ dian time to recovery for all the I individuals (we assume every. and : ker connects each dot to the half in longer than median ~ box. randomly choosing ~ half the individuals to receive I Q p the remedy and half to receive : the placebo.: distrlbunon m such a way as to I ' . 'b' . the me. The in. distrl uoon. tIve. ~ • box plot we also ~sume that no one . • . b h d b d .I. M ate a cold remedy by giving it . By ~ "box" encloses all data between defmition. of ' 1 than me. and subjects.with a "1" on it and one with a : lows selecove search of ~aro~"0"" I lar zones of a randoffilsatIon on It. w ho recovered mess 'like the sum 0 .I a vertical line inside the box to dividuals who received the: indicate the median of the data. ' dian orne lS " effect of th ' replacement firom a I.
i analysis tor for scientific notation. Substance. sessing the associations be• caliban puzzle a logic puzzle in which one is I tween two sets of variables asked to infer one or more facts : within a data set. observations one must corrupt to make the estimator take any 'value one wants. : a multivariate method for as. For in~ such linear combinations."nalysis 21 tive search may be concerned ~ relations. Each score (linear with the tail of the i combination) from either list is randomisation distribution : correlated with no other com~ bination from its list and with • breakdown point . I binations of variables (one for • canonical ~ each set) ordered by the maga canonical description of any : nitude of their correlations with statistical situation is a descrip~ each other. The fIrst such pair tion in terms of extracted vec.~ ~nly one score from the other mator is the smallest fraction of i st. a canonical correlations .I Msthmulties======= II . typically 8 bits. • calculator notation ~ • canonical correlation the symbols used by a calcula.II hreaItdownpoint I canonicalcorreUJ". combinations that show a remarkable pattern of zero cor. is determined so as to have the tors that have especially simple : maximal correlation of any ordered relationships. the breakdown point of an esti. • byte the amount of memory needed to represent one character on a computer. sequent pairs have maximal coranalysis describes the relation : relation subject to the constraint between two lists of variables ~ of being orthogonal to those in terms of two lists of linear I previously determined. The analysis ~ focuses on pairs of linear comfrom a set of given facts.
The canonical variates are ~ {straight. causal relation II • canonical variates analysis ~. multiple discrimi.22 canonical 11am. . . • catenary I analysis of variance : a curve whose equation is y = (MANOVA). The method is I tive variables is a bit blurry. finds lin I COnSIStIng of a honzontal numates an.: The distinction between catthogonal as in Principal Com.).I pended from two points forms relation analysis. The ca. sions. blue}. Polynesian}. A critical as. and canonical cor. nonica! variates then may be I • categorical variable displayed as an ordination to : a variable whose value ranges show the group centroids and ~ over categories. if one heats a rigid X/ 3 X/ 3 ) ..~ •. etc.: Native American. scatter within groups. female}. ~axIS . Calcutta.. For example. analysis I causation. such as {red. Hispanic. Some cat'uncorrelated. J' nance matrIX.ber line (xaxis) and a vertical fiormatlons of the data which I b lin ( .M. {Asian. Mricanone wants to describe among . . YSIS ear tr~s. C . I mg at e ongm (zero on each varIatIOn relatIve to the pooled . mtersect. {male. be lin) withingroup variation. curly}. I . however the vec. group differences in few dimen. tall}. num r e. Mumabai}. group variancecovariance. a rectangular coordinate svstem . capacity a method of multivariate analy.: this curve sumption is that the within. anoruca varIal .: {Delhi. causation. be thought of as a "data reduc. .thelulies . causal relation structure is similar otherwise : two variables are causally rethe pooling of th~ data over ~ lated if changes in the value of groups is not very sensible. This may . egorical variables are ordinal. . . • I groups is expressed relative to I the pooled withingroup cova. tion" method in the sense that ~ {short. maximise the among group ~um ethr e .~ egorical variables and qualitaponent Analysis. II =======.~ (a/2)(e +eA chain susnant analysis. tors of coefficients are not or.I • cartesian p ane . American. green. Caucasian. the amount of liquid that can fill sis in which the variation among : an object. closely related to multivariate ... i one cause the other to change.
tiu======= II .. MlJthemR. I center of the circle central limit theorem the central limit theorem states that the probability histograms of the sample mean and sample sum of n draws with replacement from a box of labeled tickets converge to a normal curve as the sample size n grows. an angle whose vertex is the • center (of a circle or sphere) The point from which all points on the figure are the same distance. or threedimensional without having any causal rela.II ceilingfunction I centroid 23 *================= container filled with a gas. and even if two variables I (of a chord or arc) An angle have a causal relation.. \ \ • center of a rotation the point where the two intersecting lines of a rotation meet I • centroid the point of concurrency of a triangle's three medians. that I • center of gravity causes the pressure of the gas I the mean of the coordinates of in the container to increase. their cor. the area of the probability histogram for any range of values approaches the area tmder the normal curve for the same range of values..I • central angle tion.: whose vertex is the center of a relation can be small or zero. converted to standard units. points in a figure. Two variables can be associated I two. smallest integer greater than or • central angle of a circle equal to x. I circle and whose sides pass I through the endpoints of a • ceiling function the ceiling function of x is the chord or arc. whether one. • center of a circle the point that all points in the circle are equidistant from . in the following sense: As n grows. I • I I I I I . '''' .
chance because it is approximately I error uncorrelated with every shape a random variable can be devariable when landmarks are composed into a sum of its exdistributed around lTiean posi.~ pected value and chance variations by independent noise of . or.The expected value of the ery landmark and in every di. I value is at most l/k2.. I number k>O. the chance that it does not SEs or further from its expected occur is 0%. For ability is 100%. the probability However.24=========== .".Chebychev's inequality two sets of landmark configuI for lists: For every number rations is the product of their k>O. The de. Centroid Size is used in geometric morphometries _ chance variation.variation is the same as the stanfiguration of landmarks so they dard error of the random varican be plotted as a point in I ablethe size of a "typical" difKendall"s shape space. in a list that are k SD's or furI ther from the arithmetic mean . Eve~ if an event random variables: For every is certain. the fraction of elements Centroid Sizes.I directions.ference between the random nominator of the formula for variable and its expected value the Procrustes distance between I .I chance variation is zero. by the complement I that a random variable X is k rule. centroid.of the list is at most Ijk2.cevian centroid size is the square root I a line segment extending from of the sum of squared distances a vertex of a triangle to the opof a set of landmarks from their I posite side. equivalently. the rection. . the same small variance at ev. the I square r'?Ot of the sum of the ~ variances of the landmarks about that centroid in xand y. Centroid Size is the size I standard error of the chance measure used to scale a con.certain event an event is certain if ItS prob. tion around its expected value.centroid size I Chebychev'si1UlJUR1ity II _ centroid size . it might not occur..
. the ~ degrees of freedom..f. Suppose ~ that in each trial. one less than the number of cat: egories. hypothesised : distributions of counts may be ~ approximated by a normal dis. Use of the tables ~ is notably simple. . and that ~ these probabilities are the I same in every trial. Suppose there are n in~ 25 dependent trials. The standard error of the corresponding random variable would be (2xd. so we can think of it as the probability histogram of a random variable.). : for i = 1. • chisquare statistic the chisquare statistic is used to measure the agreement between categorical data and a multinomial model that prediets the relative frequency of outcomes in each possible category.)V2. The chisquare curve is an approximation to the probability histogram of the chisquare statistic for multinomial model if the expected number of outcomes in each category is large. I Mslthema. squared distribution depends : upon only one parameter. The ex: pected number of times out~ come 1 occurs in the n trials I is n.f. the ~ expected number of times out: come i occurs is ~ • chisquared distribution ~ where expected frequencies are .f. sufficiently high..ties======= II . tribution rather than an exact : binomial distribution. defined as . k. the shape of the chisquare curve approaches the shape of the normal curve. chisquared statistic can be de: rived algebraically this is the ~ chisquared distribution which I has been computed and pub~ lis~ed historically as extensive : prmted tables. grows. and its total area is 100%.II chisquare curve I chisquared d~tion • chisquare curve the chisquare curve is a family of curves that depend on a parameter called degrees of freedom (d. the probabil...f. itythatoutcome i occurs is pi. each of . 2. so the expected value of the corresponding random variable would equal d. more generally. The chisquare curve is positive. As d.f. The balance point of the curve is d. which can result in one of k : possible outcomes. The cor~ responding distribution of the .xp1. as the chi.
tl4red statistic I circumfoYence II ~============*~=========== i • circumcenter • chord : the circumcenter of a triangle 1. a line segment whose end.~ cedure is to use an exact i biniomial test. a polygon with infinite sides • c~gory ~ • cl1'cular cone . a cone whose base IS a crrcle. ustially unhelpful or redundant . etc.~ egories is 2 an alternatIve pro. • circularity when on a search. It is calculated as a sum of : terms over the available catego. scribed circle.26 chiSlJ. points lie on a circle.). : "the circle circumscribed about I • chord of a circle : a figure.: cal outcomes depart from a. 2.~ quency based upon multIplymg : the sample size by the I h thesised probability for the : being considered (therefore 'E' will generally not I be an integer value). circling back to a previous place visited (definition. '0' represents the observed frequency for I the category and 'E' represents the corresponding expec~ed ~e.~ ries.: tions where the numbe~ of cat. web site. In situa. the line joining two pomts on i • circumcircle a curve is called a chord.. ~ given by the formula C = 2m.~ ties. CD and EF are chords of circle A. a segment whose end~oints ~ • circumference are on a circle i the distance around a circle. • chisquared statistic I this is a longestablished test statistic for measuring the extent to which a set of categori. where each term is of the i form: ((0E)2)fE. . Q Chord c "A D E F A line segment that connects 2 points on a circle. I hypothesised set of probabili. • circle the set of points on a plane at a certain distance (radius) from a certain point (ceI1ter).~ is the center of the circum.
one starts by dividing the range of v~ue~ into a set of nonov~rlap pmg ~tervals. power (polynomial). the sets are often con II . m such a way that every ~atum is contained in some class mterval. called class mtervals. quadratic. rational. • class boundary a point that is the left endpoint I • cluster analysis of one class interval. • class interval in plotting a histogram. ~ ogy. In biolinterval. if traveling along the line. I : • clockwise ~ in orientation.II circumscribed I cluster analysis where r is the radius of the circle. : ~ . exponential. the direction in . ger. or logarithmic. or : complex. The figure inside is inscribed in the circumscribed figure. which the points are named : when. I : • classify ~ to categorise something accord. • cissoid a curve with equation y2(a_ X)=X3. ing to some chosen character: istics. real. ~ the interior of the polygon is on ~ the right. inte. irrational. • class of functions family of functions such as linear. ~ ~ I I ~ : ~ . • circumscribed passing through each vertex of a figure. 27 *=============== ~ . P2 ~ • classes of numbers : family of numbers or number ~ systems such as natural. usually referring to circles circumscri bed around polygons or spheres circum _ scribed around polyhedrons. and the ~ a method of analysis that repright endpoint of another class : resents multivariate variation in I data as a series of sets.
• coefficient a coefficient. where k! (pronounced "k lying exactly on top of each factorial") is kx(kl)x(k2)x other. : which has 3 elements. of each size. Line segments that coin. ~ • combinations the number of combinations of n things taken k at a time is the I number of ways of picking a : subset ofk of the n things. From a set that has n elthe same points..b}.b. {a.c}. In multivariate data analysis. it follows that nco +nC1 +nC2 + . and the total number of subsets of a set is the sum of ~ the numbers of possible subsets . structed in a hierarchical manner and shown in the form of a treelike diagram called a dendrogram.. For exI ample. {a}. Because the number of subsets with k elements one can I form from a set with n elements is nck. usually the "function" is a variable measured over the cases of the analysis.28 corJJicient I combinatitms II ~ • coincidental lines . I {b}. +nCn = 2n.c}. The numbers nCk are cide are identical. one can I : form the 23 = 8 subsets U. {c}. b. from the set {a. A coefficient is not the same as a loading.. and the coefficients multiply these variable values before we add them up to form a score. {a. they have all I also called the Binomial coeffiI cients. {a. in general. is a number multiplying a function. x 1... c}. lines that are identical (one and : the same) ~ • collinear ~ lying on the same line. The number of such combinations is nCk = n!j(k!(n• coincide k)!). and without I regard to the order in which the elements of the subset are pickc:d. . ements one can form a total of I 2n subsets of all sizes. {b. withI out replacement.c}.
you get negative points : on the "real" axis just defined. . then press the "=" I 8 () button. mLABC + mLCBD =90" ~ LABC and LCBD are complementary angles. . num b " (t h e square roots 0 f ers e. : t emse ves accord mg to t e I rules. In _ complement rule ~ this system of notation. then push the nCm button. two angles whose measures number of combinations of m : have the sum 90°. numbers) and the yaxis is : identified with "imaginary .: . position) corresponds to the circles at different radii : addition of complex numbers I .compatible numbers : and both rescaling (enlargenumbers that can be easily ma. vented by Gauss.complementary angles (ncm ) that lets you compute the . When you a gIven set IS the collecnon of" ul" I " thi" b all elements of the set that are .compass a drawing tool used to draw ~ so that translation (shift of .set 0 . of multi. type the value of n. . inthe probability of the comple.commutative properties I properties about order of addi.. ~ Many operations on data in r MJJ. a x b = b x a" .~ tion correspond to multiplicatally.th_tics========== II . things chosen from a set of n I things.compIement f I th I b f .complex numbers tion' a + b = b + a. negative numbers). To use the button.comp e~ent 0 a su . : tion of complex numbers. the xaxis is ment of an event is 100% mi.: identified with the "real numnus the probability of the event: ~ bers" (ordinary decimals P(Ac) = 100% peA). ~ tip pomts on " s axiS hY not elements of the subset.II ~m#n~~I~~n~~~~~~~~~~~2==9 The calculator has a button ~ . first I ArSfc. braic way of coding points in : the ordinary Euclidean plane .I ment or shrinking) and rotanipulated and operated on men. then type the value of m.~ complex numbers are an algeplication.
tangrams. snap cubes. another will folcurved from the inside.g. How should I II =======MRthem4ries . • composition (of transformations) The trans.( called I concurrency). base ten blocks). formation that results when geoboards. ability situation such as flipping a coin and spinning a spinner. after another transformation. • composite transformation the composite of a first transformation S and a second transformation T is the transformation mapping a point P onto T(S(P». occurs.I suppose we are interested in the probability that some event A gon. and we learn that the I event B occurred. • concurrent I intersecting at a single point • compound eventS the point of two or more events in a prob. • concave polygon a polygon having at least one • conditional probability diagonal lying outside the poly. not convex..I pattern blocks. low. but have different radii I I • concrete materials objects to be manipulated (e. I • /I I I concentric circles circles that share the same center. I • conditional I a statement that tells if one • concave thing happens. • compose numbers put a set of numbers together to form a new number using addition or multiplication.~30~~~~~~~~~~~~~~mmW~l~i~p70b~i~ two dimensions can be proved valid more directly if they are written out as operations on complex numbers. color one transformation is applied ~ tiles.
abilities satisfy the axioms of : probability. so the conditional probability that A occurred given that B occurred is 100%. : . ~ . of A) agree with our intuition : as described at the top of this ~ paragraph. For inbetween cases. a proof of a conditional state: ment. For example. This is . On this scale. and all points on line segments that connect points on the circle to a single point (the cone's vertex) that is not coplanar with the circle. Basically.cone a solid whose surface consists of a circle and its interior. because we know that B occurred. The circle and its interior form the base of the cone. but B is not a subset of A. A must have occurred too. suppose that A and B are mutually exclusive. so the conditional probability that A occurred given that B occurred is zero. : ~ . A did not. To have a legitimate probability requires that P(S) = 100%. : ~ ~ . so we are interested in the event AB. new S be 100%. where A and B intersect. Then if we learn that B occurred. ~ . so that A must occur whenever B does. Conditional prob. : ~ . suppose that B is a subset of A. come space to B. just as ordinary ~ probabilities do. that the special cases AB = {} : (A and B are mutually exclu~ sive) and AB = B (B is a subset . pro~ vided P(B) is not zero (division ~ by zero is undefined). The radius of a cone is the radius of the base. The altitude of a cone Msthemtllti&s========== II . Note . we need to : divide by the probability of B ~ to make the probability of this . one "restricts" the outcome space S to consider only the part of S that is in B.31 *================= we update the probability of A to reflect this new knowledge? This is what the conditional probability does: it says how the additional knowledge that B occurred should affect the probability that A occurred quantitatively. the conditional probability of A given B is a number between zero and 100%. Then if B occurred. : the probability that AB hap~ pened is P(AB)fP(B). At the other extreme. so if we are restricting the out. For A to have happened given that B happened requires that AB happened.conditional proof . the deftnition of the conditional : probability of A given B.
The ?elgh~ of a cone 18 t. With fmite numbers of data the number of related distributions will be fmite. the v~ue of the I test stanstic hes WIthin this ~ range.he length of Its aln~de.two groups. a family of related distributions may be defined according to a range of hypothe~cal values of the pattern whIch the test statistic measures. of the ~istribution. Approaches ~ to providing suitable con~ ~traints.e of re: randomIsanon distributions. bet. The likelihood of the outcome value may be calculated for each distribution in II ~ the family. a related distribution may be formed by shifting all the observations in one group by a common amount where this common shift is r~garded as a continuous variable. For instance. mtet. If the line segment c~nnecnng the vertex of a cone with th~ center of its base is perpendi~ar ~o the base.values with weight corI responding exactly to an arbi: trary nominal alpha criterion II ===:========Mstht:'l'/Ul. and typically considerably smaller than the number of points of the randomisation distribution.cas. the con~ fidence interval is defmed by ' the minimum and maximum ~ values of the range of values so ~ defined. s~ch that a confidence . • confidence interval for a given rerandomisation distribution. total weight within the range of : values is regarded as an alpha ~ probabili~ ~at. Generall~ the defmition : of a confidence mterval cannot ~ be unique wi~out imposing ' further constramts. and these likelihoods . I these are discrete distributions : so there will generally be no ~ range of.~32~~~~~~~~~~~Mlmn~m~M is the line segment from the vertex to the plane of the base an~ perpendic~ar to it. include : d~g the confidence interval ~ : to mclude the whole of one tail . may be then used to defme a ~ contiguous set of values which : occupy a certain proportion of I the total unit weight of the like: lihoods integrated over all val~ ues of the test statistic. The proportion of the . or to be : centred m some sense upon the ~ outcome value. otherwise it is oblique.tU:s .ween TAILS of equal ~ weIght" In. or to be centred . then the cone 18 a nght cone.al will be unique. for the pitman permutation test to test for a scale shift between . the .
II con. .congruent clrcI es groups other than the treat..~ the treatment (if any). those differences beI image of the other under a retween the groups are said to be ~ flection or composite of reflecconfounded with the effect of : tions. likely to affect observational the interval that will ~esult on~e : studies and experiments that data are collected wIll contam I are not randomised Confoundthe corresponding parame~er.congruent angles : two or more angles that have • confounding when the differences between ~ the same measure.congruent figures ~ two figures where one is the ment.confidence level : confounded with the effect of the confidence level of a confi. For exuniqueness is therefore not gen. vals again and again from inde.fideme level I congruentfiDU1'eS.~ smoking. those factors would be .: . Confounding is quite dence interval is the chance that . . : questioned whether differences ~ between individuals that led . etc.: randomisation. equal. some to smoke and others not : to (rather than the act of smok~ ing itself) were responsible for I the observed difference in the ~ frequencies with which smok: ers and nonsmokers contract I various illnesses. If that were ~ the case. I sponse that are not distinguishable from the effect of the treat.~ pendent tiata the longterm .congruent limit of the fr. If ~ ing tends to be d~creased by one computes confidence mter. • the treatment and control I: .ction of intervals ~ equilateral. . and the problem of non. prominent statisticians eratly solvable. MIIthemtlnes======= \I . shape.: same radius.) the confidence level. exactly the that contain the parameter is I same (size. ample.I two or more circles that with the ment produce differences in re. 33 level.
 :... .. • consecutive sides (of a polygon). An ellipse..: .. the Procrustes mean shape is computed to minimise the sum of squared Procrustes distances from the the consensus landmarks to those of the sample.... I II I • consecutive vertices (of a polygon or polyhedron).~ . . .Xo = Vat + tit . ~ ~  I .. • conic section the cross section of a right circular cone cut by a plane.. usually made as a result of inductive reasoning.Xo = (tJo + tI)t 2 II =====. parabola.vat 2 2 1 1 x ... two vertices that are connected by a side or edge.. I • I I I I ='= I ( I . •...I ..va) 2 x . Often a consensus configuration is ·computed to optimize some measure of fit to the full sample : in particular.. • conjecture a guess.'...\ .tpolygOllS I constantofRnetpmtUm • congruent polygons two or more polygons with the exact same size and shape. including the point and the region..... =:. example: "0 1 (v . I I I • consequent the second or "then" part of a conditional statement...Xo = vot + ....l_tt.../ ) '/ i I ~ \\ /  l ' 1\ \  I \ '/ I • conic solid the set of points between a point (the venex) and a noncoplan. / \ ! 1\ .Xa = .. \ " . • congruent segments two or more segments that have the same measure or length. constant of an equation the term that has no variable in an equation... and hyperbola are conic sections. two sides that have a common vertex...... consensus configuration a single set of landmarks intended to represent the central tendency of an observed sample for the production of superimpositions.===MsJthmuJnes ....\ I \ f. two angles that have a side of the polygon as a common side. or some other morphometric purpose...vat + vt I 1 1 2 2 1 x ..:lI region (the base). of a weight matrix. • consecutive angles (of a polygon). :\ ..~34~~~~~~~~~~C~ongruen=..t 2 t I • I I x ..
cesses is np = 10. area under the normal curve : between 0.~·ty=~ • constant rate of change set . and a binomial random variable : is discrete.20 standard units. and the stan~ dard error is (np( Ip» lh = 6lh .~te~oJi~'ch~'Q.6)15 = ~ 16%. ~ The expected number of suc.20 is ~ about 15. we need to "smear ~ out" the binomial probability ' over an appropriate range.20 and +0.5 10)/2. • construction a precise way of drawing which allows only 2 tools: the straightedge and the compass • continuity correction in using the normal approximation to the binomial probability histogram. In a similar way. • construct create a figure using only a straight edge and compass. 10 : 1/2 successes.5 10)/2. lower endpoint of the range. The ~ upper endpoint of the range. The true bino. under the normal curve at the : point 10 successes transformed ~ to standard units.45 ~ = +0.' we get zero: ~ the area under a point is always . For example.45 '. mial probability IS : 25CI0x (0. 35 .II ~consts~~n~tffl.~~~e~Iconti~~·n~Uf. transformed to standard units. The number of successes in this scenario has a binomial distribution with parameters n = 25 and p = 40%. The . suppose we seek to approximate the chance of 10 successes in 25 independent trials.20 standard units. We get a better approxi: mation by considering 10 suc~ cesses to be the range from 9 . zero. Because ~ the normal curve is continuous . = 0.45. . = 2. 9 : 1/2 successes. so this is exactly right for the : binomial distribution.of data or table of values in which ~e amount of the dependent vanable changes by a constant (fixed) value as the value of the independent variable changes by a constant value. If we consider the area . is (10. The only : possible number of successes ~ between 91/2 and 10 1/2 is 10. The . each with probability p = 40% of success. if we ' seek the normal approximation : to the probability that a bino . is (9.8%.4)10x (0. This is clearest if we are seeking the chance of a particular number of successes. one can get more accurate answers by fmding the area under the normal curve corresponding to halfintegers. 1/2 to 101/2 successes.
inclusive. II mial random variable is • the ~ for any possible ~a!ue of ~e range from i successes to k suc.. c~n:espon~ value of the area under the normal curve : the statIStiC m question. inclUde temperature. ~ the oppo~lte of dilatlo~ a ~. trans. by a number betwec:n zero and continuous statistic. probability distributi?n function cesses. based upon ~ one. ure resulttng from multlplymg • continuous distribution ~ all dimensions of a given figure a probability distribution of a .36 I cuntinUIIIIS tlistriburion I cuntmaitm . A random . we ..lone can never measure a conformed to standard units. and the chance to standard units. I • continuous variable transformed to standar~ . I .. exact the area under ¢e normal curve : height exact age (including corresponding to the range i + 1/ ~ parts of a second). cumulative probab~ty there IS cesses. unde. from iI/2 to k+ 1/2 successes. ~ a quantitative variable. is conIf we seek the probability of : tinuous if its set of pOSSible valmore than i successes and fewer ~ ues is uncountable. we should find the area: discrete variables. If we se~~ the : that it takes any particular value probability of at least I but ~ is zero (in symbols. . such that : I . transformed to standard : is a continuous funCtIon (a funcunits. Including or excluding the ~ tion .r the normal curve corre. • halfinteger ranges at the ends I of the interval in this manner is : • contrac~on . = 0 for every real number x).: variable X is also called conttnusponding to the range i + 1/2 to I ous if its set of possible values k+ 1/2 successes. should fmd the area under the : A random variable is continunormal curve corresponding to ~ ous if and only if its cumulative the range iI/2 to kI/2 suc. ifP(X = x) fewer than k successes. called the continuity correction. In practice.units. we should find . transformed ~ is uncountable. If we : tinuous variable to infinite preseek the probability of more ~ cision so continuous variables than i but no more than k suc. an algebraic formula.with no jumps).. are so~etimes approximated by cesses. 2 to kI/2 successes. we should find ~ an exact. Examples than k successes.
: . and to . by tossing a : com. There are many methods that try to control for variables. others use assumptions about the nature of the effects of the variables to ~ to model the effect mathematically. convenience.37 II contm. I ~ : . method of comparison to evalu: ate the effect of a treatment by ~ comparing treated subjects I with a control group. dom.convenience sample ~ a sample drawn because of its . I might ~ take a sample of opinions in . using regressl0n. for example.positi1le I cunwnien&e sample ================*================ . not receive the treatment. Samples of convenience ~ are not typically representative. I : . a controlled experiment. The contrapositive is logically equivalent to the original proposition. so it will not confound with the treatment. . randomised I • : expenment I a controlled experiment in ~ which the assignment of sub: jects to the treatment group or ~ control group is done at ran. Delhi (where I live) by just ask: ing my 10 nearest neighbors. I : That would be a sample of conI venience. and it is not typically possible : to quantify how unrepresentaI MMIIenIaeics======____. Some are based on matching individuals between treatment and control. control for a possible confound: ing variable. . for example. not a probability : sample.controlled experiment ~ an experiment that uses the .controlled.control group the subjects in a controlled experiment who do not receive the treatment. For example. to whom no treatment is given. and would be unlikely : to be representative of all of ~ Delhi. then the contrapositive of the proposition (p IMPLIES q) is the proposition «NOT q) IMPLIES (NOT p) ). control there are at least three senses of "control" in statistics: a member of the control group. who do .11 . _ control for a variable to control for a variable is to try to separate its effect from the treatment effect.contrapositive if p and q are two logical propositions.
x2. . .convex set If such a number x eXISts. : _of __ of _. P( 1X2 ~ segment (if any part of the segXI < e).. the statement formed by exchanging the "if" and "then" parts 0 f an "if.. x3 ... x set lie entirely in the set..... converges if there is a number x such that for any number E>O. it's converges to 100%. . tive results based on samples of ~ 2. there is a number k (which can depend on E) such that Ixj xl < E whenever j > k.th" statement . ...:: _ converge convergence a sequence ~f numbers xl. 1>. : 2. of the proposition (p IMPLIES ~~=.convergence in probability ~ are three things one can do to a sequence of random variables i see if a figure is convex look for Xl. .. i_convex polygon : a polygon having no diagonal ~ lying outside the polygon. the sequence of : points within the figure with a numbers P( IXI XI < e). X2. one system ~o the other (e. v.38 .. .."" .. P( IX3 XI < e). It 18 f hich all eg called the limit of the sequence ~ a set 0 pomts om w fSth 1 x2 3 i ments connectmg pomts 0 e x.!!!!!·c~O'1¥I1i~m!!!!!~!!!!!mee~!!!!!1coo!!!!!~"~rJt~~na~te II. . T_.# 'Ii>! 00.. I .converse 1. if P and q are two logical i propositions.:~ q) is the proposition (q IM~ i i PLIES p).' I ! COl ~.en : _ coordinate ~ a number that identifies (or I helps to identify) a point on a : number line (or on a plane. . .. .: a. > .. . then the converse convenience will be. ~~~~WD~e.. or I . . ':" ~ : : : . (of a conditional statement). 1. . 0 0 0 0 iconversion factor : relationship between two umts ~ from different systems of meai surement used to convert from ... and connect any two number E>O. . extend the segments probability if there is a random ~ (they shouldn't enter the figvariable X such that for any jure). _001 ... . : concave).. : m space) I . There .~4 centimeters corresponds to ~ 1 mch).:.g. X3 converges in : "dents".:=. • 0 0 • 0 : 0 0 0 0 0 I .. I ment lies outside the figure. c.8UJIa!i1C9 ..
n) =1. _ coordinate system : from the first set of ordered pairs used to 10. a onedi~ mensional graph. : .~ _ corollary to a theorem . Frequently the word M/lthemR.coordinatised.coordinate geometry the study of geometrically rep.: sense. ber and vice versa. .correI abon the origin.coordinates a set of parameters that locate a point in some geometrical I space.coordinate plane a plane in which every point is I _ coplanar identified with exactly 1 num. coordinates in a geometric resenting ordered pairs of num.. lines through one single point. line ~ a line on which every point is . . a twodi. The elements of any I relation between two or more . locate a point on a plane or in physical space by I projection onto perpendicular I . variables. The distance I between 2 parts on a : coordinatised line is the abso~ lute value of the difference of I their coordinates. a proof using coordinate geom.~ lying in the same plane.. I bers : . .39 ~ vector may be thought of as . a theorem that is easily proved etry.coprime mensional graph : integers m and n are coprime if _ coordinate proof ~ gcd(m. identified with exactly 1 num: her and vice versa.I p cate an object or point on the I twodimensional plane. for instance. Cartesian coordinates.riu======== II .
there are uncorrelated. . a set is countable if its elements ments of the transformed : can be.. + 1.This . { . Spearman. ... also Kendall.Y).too~ tha~ lets y?U I {straiWtt.I in any right triangle containing cient is always between 1 and I the angle. The correlation coeffi. The correlation co.. tion coefficient of a list of . If a . "I • countable set Multiply correspopding ele. etc. . This page . correlations I if two parallel lines are cut by a which measure other aspects of ~ transversal.. correlated without having any correlation coefficient IS + 1 or causal relationship. blue} list of products. corresponding die relation between two vari. • coterminal angles pairs of measurements (X... curly}. For eXaniple the sets efficient is the mean of. and two 1 when all values fall on a variables can have a causal restraight line. }' ~. integers. contams a . " :1" " 2 . . gre~n. . that {O~ 7 3} {red.angles are translations of each ables. To compute the correla. product of the standard ~ Two variables can be strongly deviations. is used for Pearson's producrI • • correlation coefficient • cosine the correlation coefficif!nt r is I (of an acute angle) The ratio of a measure of how nearly a the length of the adjacent side scatterplot falls on a straight I to the length of the hypotenuse line. However.Sy. other along the transversal . not parallel to ei. : two angles that have the same first transform X and Y indi.put in one~toone correpairs to get a single list of ~ spondence with a subset of the numbers. it is un. want.correlation moment correlation which is I a measure of linear association the covariance divided by the between two (ordered) lists.~" terminal side vidually into standard units.I lations hip and yet be ther axis. set is not countable.rxy=Sxy / Sx. and the set of generate" bIVarIate data WIth : all fractions are countable.. • corresponding angles tetrachoric. I ' · any correIanon coeffiICIent YQU . '2 0 1 .
a I I ~ • cover . I Before the data are collected. lllt. the relation between invariants and covariants is a rotation by 90 degrees in the shapecoordinate plane.=======II .tha.. C<:MI:I'iAntl"lft.lrlcTII'rtSf. For more than three landmarks.II counterclockwise I criticalveUue 41 ~~~======~*==~=========== countable. trees and lists.._orreI4UM ..rflcllll'4 . For transformations of triangles. but : a full plane (four landmarks) or • counterclockwise in orientation. • covariant a covariant of a particular shape change is a shape variable whose gradient vector as a function of changes in any complete set of shape coordinates lies precisely along the change in question. • critical value . . (aC'lt. aka contradiction • counting techniques a variety of methods used to determine the total possible outcomes. dence intervals is the chance : that the procedure produces an ~ interval that covers the truth. : the chance that the confidence ~ interval will contain the paramI eter value is the coverage prob~ ability. the critical value in an hypoth: esis test is the value of the test I ~ Msthernades. • coverage probability : the coverage probability of a ~ procedure for making confi.~ rl'(HTl ")tIJvJ1(1U7'PolnbLtIC. a confidence interval is said to : cover if the interval contains the ~ true value ofl the parameter.~J A • counterexample a situation in a conditional for which the antecedent is true..¢qUCOtr~!lI. including the multiplication principle.. typically in a probability situation.r rr'. which equals the confi: dence level after the data are I collected and the confidence in~ terval is actually computed. The set of all real ~ given transformation has only numbers is uncountable.WJst. but the conditional is false. · ' if travelling on the line. lone direction of covariants... the direction in ~ hyperplane (five or more landwhich points are named when I marks) of invariants. the interior of the figure is on the left side.
sectional study would be likely : to confound the effect of age ~ with the effect of the mores the ..: 1 . it looks at a crosssection of a . Each letter represents a unique digit.I ... ed·th . : ~ . the ch ance that the rand om IS . For example.a:' ably r~s WI a very ~ler. The critical value is set so that the probability that the test statistic is beyond the critical value is at most equal to the significance level if the null hypothesis be true. pares different individuals to ~ each other at the same time.42 crossseaional study I Ciunulati. b bili'ty um a ve . 1 than ual varia e IS ess or eq to fun' f In bols X. population.ero a D' m'bU u'on F unc0'on IS . sym .. then F(x) = P( X < = x).. CUlty to e agmg process. if F is the cdf of the random variable X. th' C f. • cube a solid figure bounded by 6 congruent squares. The differences between those individuals can cQn. ..a: • . as a etlon 0 x. • a crosssectlonal study com. bl etlon 0 a ran om varia e . ~ II statistic beyond which we would reject the null hypothesis. in trying : to determine the effect of age ~ on sexual promiscuity.: attn ute wuerences m proIDlS. subjects were taught as children:: .. theoIder mdiVI'duals were prob. r I. ... bl . . 'b . ub' equa0'on • C IC I oml 'al ti f d a po equa on 0 egree .. ~ntthanatt1tuthde towards proIDlScu. (CDF) the cumulative distribution fun' f d . Thus It would be Imprudent to .. a cross. Probability Distribution Pmu:tUm. The cumulative distribution function must tend to zero as x . C u1 u' 'D. ~ : I : • crosssectional study I . Ity e younger sub~ects . · . indicated arithmetical operation : I has some or all of its digits replaced by letters or symbols and where the restoration of the original digits is required.1:... : • cryptarithm ~ a number puzzle in which an .~ found with the effect being ex.1:. plored.~ .rudinal tud I ong! s y.
cylinder . . there IS.s~ce. : _ cyclic quadrilateral . when laid . . : including llle region an Its . . etc.II :. : a solid whose surface COnsISts of ~ all points on two circl~ in ~o I parallel planes. a space wlth coordinates and a distance f~ction such that the area of Circles.curved ~pace. Kendall's shape space. then the : cylinder is a right cylinder. The I : height of a cylinder is ~e ~ength ~ of an altitude. along WIth pomts : in their interiors (the bases of the ~ cylinder)... : The axis of the cylinder is the line ~ segment that joins the centers of I the bases.cycle ~ age (of a periodic curve) One se~tion of a curve that.cylindric s?lid ~ the set o~ pomts ~~n a re. For mstaI?ce. the usual iI?tui tions about what "straight lines" can be expected to do .fue approaches minus inftnity.. I •• • . and perpendicu: lar to the planes of the bases. are not pro~ortional to the ap. gion ~d I~ trans~nondm. thenF(y) >= F(x). just like there IS a pomt on the surface of the earth as far as possible from where you now sit. It is a positive function. g.cylindric surface out repeatedly endtoend. e. .. oth~ erwise. ... volume of spheres. a polygon whose vernces lie on ~ a circle.cyclic polygon . The radius of the cyl: inder is the radius of a base. In curved spaces..lffi ~lhemsti&s======= II . 1~tur. will be fa~lty. *=========!!!!!!4!!!!!!3 ~ . : the union of the bases and the I : lateral surface forms the entire curve. and all points o~ line I segments joining the two arcles. it's oblique. If the axIS IS per.. and must tend to unity as x approaches inftnity. another t?at I! "as far fronl ~t as 1?osslble. . corresponding to every trlangular shape i? Kendall's sha~e space.. : . pendicular to the basc:s... An ~ altitude of a cylinder is a line segI ment between. . scribed in a circle.propnate power of the radIUs. The cumulative distribution function completely characterises the probability distribution of a random variable. and increases monotonically: ify > x. ~ a quadrilateral that can be m.
It is typically as. and m which each definition uses only terms that have been defmed previously in the system. than the nominal alpha criterion : level leads to a decision of sta. and clubs.~ come value of alpha with a . . For example. ing that the test statistic has a : value other than its (null) ~ hypothesised value. dia. hearts. An outcome ~ value smaller (more extreme) . The thirteen . a digitiser can be used to supply information of other sorts. cards of each suit are {ace 2 3 : 4? 5. • deficient coordinate in addition to landmark locations.~ monds. 13 each of ~ four suits: spades.7. . I queen. a standard deck of playing cards : contains 52 cards.~ tistical significance of the find. it is equally likely . .. .44 =================* dartI dejieientcotmlintJte I . I • decision rule : a rule for comparing the out. The null model of independent Gauss~ noise does not apply to posItion along the tangent ?e II========~~Mstherntmes . to be in each possible ordering. in which each theorem can be proved by deducti~e reasoning using only the premIses ~nd ~revious theorems. 10.6. nominal alpha criterion level: (such as 0. q~n: ~ king}.: sumed that if a deck of cards is ~ shuflled well. break up numbers into addends reasonmg. • decagon ~ • deductive reasoning ~ reasoning accepted as logical a tensided polygon • decimal number .8.9. ~ data .05). inform~tlon ~ The~e ~ are 52! (52 factorial) possible orderings.to encode part of the inf~rma~o~ about a curving arc by Identifying the spot at which the ar~ lies farthest from some other unage structure (perhaps another such curving arc). a point can used . • decompose numbers used as a basis for . : and proven facts. from agreedupon assumptions a number written to the base 10. The face cards are {jack. or factors. king}.dart a concave kite. jack.deductive system an arrangement of premises and theorems. ~ • deck of cards .
easure for angles and ~ a density scale. A right ~ each bin .~ dent or explanatory variable). . me~urement that IS a comneeded for its computation . bl s bolized b 90°.density . . the ratio of the mass of an ob. angle measures 90 degrees. For ex. This is called a unit of m. . .: other variable (the the mdepengrees of freedom" of s?me sta. .I "dense" the observations are in tation around a circle." The usual model of independent G~ussi~ ~oise is inapplicable m pnnclple for such points.~ pendent variable. denominator. an~ s~ that Cartesian coordinate IS defi cient.dense line .definition : ject to its volume. ~ . has units of percent per unit of .I plains the meaning of a word : . The : I MMhemsnes=========== II .degree : the horizontal axis. . ~ missing observation can be ..depen nt vana e . . ym y : in regression. ~ whose 'values are supposed to given a set of parameters ~stt.deficient number : the line that contains the shona positive integer that is lar~~r ~ est path between two points than the sum of its proper diVlsors.density scale ~ the vertical axis of a histogram or phrase.. de . the de. computed as. _ denominator : in the fraction xly. a statement that clarifies or ex.45· II MJkimtnumbwl derWed~~~~~~~~~~ direction of the curve that is digitised in this ~ay. it measures how arcs equivalent to 1/360 of ro..degrees of freedom . bmatton of two other measuregiven the sample mean. be explained by chang~ in the mated from the data. Usually one regresses th~ dedent observanon~ ~equ'tred to : pendent variable on the mdecompute the stattsttc. x is called the ~ numerator and y is called the . tistic is the n~ber of m~epen. grees ~f freedom because only : . the vanable . ample the variance has n1 de.derived measurem~nt n1 of the observations are ~ a..
the differences be. .~ are points <. a CIrcle. In comput. measuring difference in 10With numbers both in accu. • develop : the segment whose endpoints to be ~volved in reasoning. ex.~ • diameter of a circle (or . Emracy and speed. ~ some chosen properties or cri.at part of calculus that deals ing the SD. : polygon or polyhedron. cation between two samples ~ wi~ inte~alscale data. In to know. one fmds the rms : ~I~ the opeation of differenof the deviations from the ~ tIatIon of functions. based upon the mean of the data..=========. teria. For example.~ of that segment I • difference of means lationships. mean.• ameter to know.~ of the circle. become skillful in working . a test statistic of intuitive appeal to. or make it possible ~ the. sphere) termine a triangle. b etween a d atum and some : reference value. I di • determine .. • differentiate tween the individual data and . : that passes through the center three Sides determine a tri _ . . . typically the ~ • differential calculus mean of the data. three angles do not de. to distinguish from other mem: bers of a class. ploymg this test statistic in an • deviation I : exact test defmes the pitman a deviation is the difference I permutation tests(l or 2). longest ~ord of ~ figure.. • develop fluency . the.>n a circle (or sphere) ploratlon. I a line segment connecting two • describe : nonconsecutive vertices of a I to explain orally or in writing. 46 describe I dijfermtUJte II ments such as speed in miles per ~ • diagonal hour. a diameter IS a chord of a fi?ure. center of the manipulatives or sketches to gain : CIrcle as Its mldpomt. conjecturing. all the characteristics . th. angle. the length understanding of concepts or re. using I ~t co~tams. for.
a translation goes. 4. and height of . 2. : randomisation distributions. based upon ~ algebraic formula or upon reI randomisation or upon actual ~ data. • dimensions ~ and also of test statistics which the width. 9.II tIi. meeting in space. : or negative I • digit in the decimal system. the graph is a straight line through the origin. I.dihedral angle : a circle together with its intethe angle formed by two planes ~ rior.: associated with possible values ric figure by a scale factor rela. in which the cumulative c : probability increases in non. : . situation is characteristic of dilation). one of the numbers 0. 8. 5. 7.direction of a translation ~ the compass direction in which . an • direct variation a variation of the type y = kx.dis.crete distribution . 6.~ of the statistic in question. ~ _ disc . 3. length. _ diophantine equation : a line made of dots with space an equation that is to be solved ~ in between their centers. in integers. a probability distribution of : some statistic. are essentially discrete. the way a number goes positive represent other digits. E .gimetie I discrete line 47 • digimetic ~ .direction a cryptarithm in which digits . ing to noninfmitesmal weight enlarges or reduces a geomet. ~ . a plane or space figure ~ _ discrete line /. This tive to a point (the center of the .dilation ~ infmitesmal steps corresponda nonrigid transformation that .
. . . it : should never be used without a ~ prefixed adjective to qualify it.g." C f . . i di·· tuall I : . i. I • distance . con. sJOlnt or mu Y exc uI sive events . idissection : the result of dividing a figure ~ into pieces. at 15. 1. _ disctiminant function a discriminant function is used to assign an. the distance between points : A and B is written as AB ~ 2. I .. the volume of fluid that rises above the original fluid line ~ when a solid object is placed i into the fluid . Typical examples of discrete variables are variables whose possible values are a subset of . d d peopIe m a aJ. ·bI·th I f th th th . i Mahalanobis distance. . be th cunty num rs. ~~======. screte random variable is one whose set ·bl of POSdSI e va l '1~ cdi~untabieif· . e number of · f:~_:1. D15crete " hunky. two events are ~omt or muI tuall . to the nearest year.. .. the mtegers. dis· . yexc USlve e occurrence . occurrence 0 eo er. such as SOCial Se. two events : are disjoint if their intersection I · th IS e empty set. I ifth ey can' t both happen at once I (ifth ha . 0 f one IS mcompan e WI the .48 . Euclidean distance. Linear discriminant functions take a vector of observations from a specimen and multiplies it by a vector of coefficients to produce score which can be used to i classify the specimen as belonging to one or another predefined ~ group.e.. ey) Eq~ nal° outltcome m comI mon .ubes A ran om vana le 15 screte · ula . if it is piecewise constant and only increases by jumps. ~ e.IlllY. bl vana es are bl A di . ages roun e . ~ a • discriminant analysis a broad class of methods concemed with the development of rules for assigning unclassified objects/specimens to previously defmed groups. . this term has several meanI ings in morphometrics.discrete variable a qUantitative variable whose set of possible values is countable. b and 0 nly if Its cum nve pro ability distribution function is a stairstep function.observation to one of a set of groups. etc. ~ • displacement . c ttnuous vana e. ifth . wv en y.
The distnbutlon func.distribution function. and is unity for large enough values of x. the empirical distribution function evaluated at y is at least as large as the empirical distribution function evaluated at x. the height of the . the probability dis~ tribution of a random vari. The ~idth of the sta~rs ~ _ distributive property IS the sp~cmg be~een adJa. : . t han or equaI to x. 3.distance between 2 parallel ~ lines the length of a perpendicular I .distributive law function is an uneven set of : the formula a(x+y)=ax+ay. the length of the translation vector between a figure and its image.~ thepropertywhichrelatesmulcent data.: tion is zero for small enough (negative) values of x.165_ . : characterised by its probabil~ ity distribution function. cal distribution function. (of a translation). f' h b bT re al vaI ue 0 f x. or fit e pro a I tty . ~tairs. ~ .. a(b + c) = a x + a data have e~ac~ly ~he same I x c. segment between them : . 'b . : Similarly. A pIot 0 f the empirical distribution ~ . .Procrustes distance. It increases monotonically: if y > x. taxonomic distance. able is completely . : bution" is used as a synonym of numencal data IS. t he firactIon 0 . . ues are distributed over the : real numbers. Sometimes the word "distridistribution function of a set . empirical the empirical (cumulative) . b . tiplication and addition' the stairs depends on how many ~ formula. for each I tior the empirIc al distn utIon . It is completely ~ characterised by the empiriAF02. f ' observations that are less ~ d~nc~blOn. value. Istn utIon unctIon. : ~ .distribution ~ the distribu~ion of a set of nuI merical data is how their val.
I diVIsor. neither the subjects nor the people evaluating the subjects knows who is in the treatment group and who is in the control group. pentagon. : • divisor ~ in the expression "a divided by . A : regular dodecahedron is a ~ regular polyhedron with 12 I faces. ers and protractors. : ~ . connecting each set with line segments. : • dodecagon. b". • d omam . rul~ . • dot a description of a point in which the point has a definite size • d ouble I" graph s me graphs in which two sets of data are graphed at the same time.50 ~~~~~~~~=* . a freehand picture using any : tool. b". . ~ a twelvesided polygon I • dodecahedron . • drawing function is defined. a solid figure with 12 faces. • doubleblind. : ~ . . • domino two congruent squares joined along an edge. mon x the domamof a fun' fi()'IS I• the set of x values for which the . diPitlend I drawing II • dividend ~ in the expression "a divided by . a is the divident and b is the : . doubleblind experiment in a doubleblind experiment. : I .a is the divident and b is the : I VIsor. di .draw : to create a figure using num" ~ bered scales on tools such as . . Each face is a regular. . This mitigates the placeho effect and guards against conscious and unconscious prejudice for or against the treatment on the part of the evaluators.
I . .~ principal components for outages of groups of individual~. I . They are I . At the dividuals. : The formula for thinplate ~ splines on landmarks can be dual I extended to encompass data (of a tessellation). : tian fraction. tween pairs of outlines. EcologI.edge . : which have the propernes 0 f ~ principal component vectors .. the new ~es : about edgels as well.Egyptian fraction system . through a landmark. : end one gets "eigenshapes. I : .duodecimal number : . x is an integer is called an Egypbase 12. edgel specifies rotation of a di: rection through a landmark. .ecological correlation : . sellation formed by connectmg : intended eventually to circum(with line segments) the centers ~ vent any need for deficient coof polygons with a cOI?mon .II dual I eigenshapes 51 *~============= ~ . or both.eigenshapes the correlation between aver. An eigenshape instead of individuals. : phometric analysis.: analysis begins with the seleccal correlation can be mislead. An . (of a solid) The intersection of : two faces. ~ extension along a direction . ordinates in multivariate moredge in another tessellanon.edgel I an extension of the notion of ~ landmark to include partial in: formation about a curve ~ through the landmark. line data." I . ~ a number of the form l/x where the system of numeration with .~ tion of a distance function being about the association of in.
a geometry in which there are : no parallel lines. I This law: which is in fact an : assu'llp. • eigenvectors in the equation given to define eigenvalues. e. • (uncorrelated. so . • elliptic geometry ."tII'lIJIes II • ellipse a plane figure whose equation is x2ja2+y2jb2=1. describing the· ~ sample in decreasing order of . t~e first elge.. I elliptic Fourier analysis a type of outline analysis in which differences in x and y (and p~ssibly z) coordinates of an outline are fit separately as a function of arc length by Fourier analysis. EtE=I and EEt~I) . and logarithmic functions. exponential functions.that the ~ scores for each speclme~ of ~ the sample can be combll~ed . . ~ • empirical law of averages ~ the Empirical Law of Aver. E contains the eigenvectors. • ellipsoid a solid figure whose equation is x2ja2+y2jb2+z2jc2=1. ages lies at the base of the fre: quency theory of probability. matrix (i. and so on. I states that if one repeats a random experiment over and lover.ion abo~t how'the world works rather than a I mathematical 'or physical law. • elementary function one of the functions: rational functions. trigonometric functions.greatest vanance. E is an orthonormal. When sorted by descen?IDg elgenv~ues. In the common data analysis case.52 7 :enPeCtors I empirietdlmPof.nve~tor IS that bnear combmatl0n of var~ables that has the .. : Eigenshapes apply to curves ~ as relative warps apply to ~ landmark shape. The second elgenvector is t~e linear combination of vanables that has the greatest variance of such combinations orthogonal to tl}e first.. variance) and also are outline: shapes themselves. ~ . to produce a new outhne : shape that approximates it in I some possibly useful way. the fraction of trials that result in a given outcome converges to I a limit as the number of trials grows without bound. independently and under I "identical" conditions.
one ~ circle as it rolls around another "must decide whether to in. . : ~ . • ends of a kite the common vertices of the equilateral sides of a kite .~ • equiangular polygon ~erval.II empty set I equidisttmt 53 • empty set ~ • enumerable set the empty set. w~ch mcludes both of . denoted {} or 0. circle. • epicyc10id : the locus or path of a point on a • endpoint convention in plotting a histogram. . which includes both I of its endpoints. I The two standard endpoint conventions are to include the left endpoint of all ~lass inter. equichordal point . a countable set. a polygon all of whose interior ItS endpomts. clude a datum that lies at a Eptcycloid class boundary with the class interval to the left or the right of the boundary. : angles are equal. • equidistant equally distant. The rule for making this assignment is called an endpoint convention. . is the set that has no members.enneagon a ninesided polygon ~. . • end~oihts . excep~ fo~ the leftmost in.I vals and exclude the right.~ • equianglular clude the right endpoint of all : having angles of the same meaclass intervals and exclude the I sure left. : I ~ a point inside a closed convex curve in the plane is called an equichordal point if all chords through that point have the same length. exCoustle Cuns (wrt horizontal reus) cept for the rightmost class interval. and to in. the pomts at the ends of a line segment or arc..
escribed circle an escribed circle of a triangle is a circle tangent to one side of the triangle and to the extensions of the other sides. An estimator is a random variable. or (most simply) the total of scores in one designated group. is the difference of means. In such a situation a randomisation test performed on either of these test statistics will necessarily have the same outcome in terms of alpha. with corresponding savings in amount of computation required. • equilateral triangle a triangle whose sides are equal in length I I I I I I I in place of a test upon the descriptively more valid one. but simpler equivalent test statistics include the mean for one designated group. In this situation. If one of the statistics is of good descriptive validity whereas the other is simpler to compute. I : ~ . then a raridomisation test upon the simpler statistic may be used I I I • I : I . II ~=====~Msthemstir. error tolerance the value allowable above and below a number or its approximation.s . as defined for the Pitman permutation test. tistics may be interrelated in a manner which is provably monotonic irrespective of the data. • estimatot: an estimator is a rule for "guessing" the value of a population parameter based on a random sample from the population. because its value depends on I • • equivalent test statistic within a randomisation set. it is possible that two different sta.tqr II • equilateral equal in length • equilateral polygon a polygon all of whose sides are equal.54 elJuilateml I I I esti'flUl. the descriptively valid statistic. An example of such equivalent test statistics occurs for the situation of comparison of levels of a single intervalscale variable between two groups.
euclidean distance matrix analysis EDMA. : ~ . Named after Swiss mathematician and physicist Leonhard Euler. ber of arcs . orthocenter. ~ _ Euler's constant I the limit of the series 1/1 + 1/ ~ 2+1/3+ . +1/nln n as n goes : to infInity.. • euclidean space a space where distances between two points are defmed as Euclidean distances in some system of coordinates. • euler line the line through a triangle'S circumcenter.Euler's formula for poly· ~ hedrons octahedron I an eightsided polyhedron. The : regular octahedron is one of the ~ Platonic solids. and the circumcenter of the tri: angle. I : .point circIc of ABC. and then comparing those averages between samples by computing the ratios of corresponding mean distances.t==========5=5 which particular sample is obtained. which is random. I _ even function .II euclitleandistancema.. ."". ~ . When the random variable X is observed. that determines whether or not A occurs: if the . the line segment containing the : centroid of a triangle. : ~ . A canonical example of an estimator is the sample mean. A method for the statistical analysis of full matrices of all interlandmark distances. a function f(x) is called an even : function if f(x) =f( x) for all x. A Euler or nine.t'l'ixanalysis I . : I : ~ . Its value is approxi~ mately 0. whose ~ endpoints are the orthocenter . I : .even nwnber an integer that is divisible by 2. _ event an event is a subset of outcome space. which is an estimator of the population mean. averaging elementwise within samples.even node ~ a node that has an even num. An event determined by a random variable is an event of the form A=(X is in A).Euler segment .577216 . and centroid.
Notable examples are the exact binomial test. I • excircle ~ an escribed circle of a triangle.. ..56 ~~~~~~~=. • exact number a numerical result that has not been rounded or estimated. cur. I sampling which generates every I randomisation exactly once. fisher test. known to generate every randomisation. promulgation of the ideas of re: randomisation statistics. but not both exhaustive rerandomisation : a series of samples from a ~ randomisation set which is .the characteristic of a rerandomisation test based upon exhaustive rerandomisation. that is.. A does not oc. present glossary. rather than through continuous approximations which are used especially where sample sizes are substantial. In particular. A2. I • ~ . . value of X happens to be in A. if not. : . and various nonparametric tests based upon ranked data... } is exhaustive if at least one of them must occur. if S = Al U A2 U A3 U . • exactstats : this is the name of the academic ~ initiative which produced this . I • exact binomial test a statistical test referring to the binomial distribution in its exact algebraic form. the Pitman permutation tests ( 1 and 2). exactstats is : a closed email based discussion ~ • exhaustive a collection of events {AI. where S is the outcome space. I : I ~ • exclusive OR ~ one or the other. ~ group for the development and A occurs.excenter ~ the center of an excircle. that the value of alpha will be fixed irrespective of any random sampling of randomisations or uponanydistributionalassumptions. A3. • exact test .
. .. the s~pling distribuvalue of their sum is the sum of t ~on of the chisquared statistic their expected values (E(X +Y) : IS approximated increasingly = E(X) + E(Y». t • noted EX or E(X). normalIsed by the expected think of the expected value of a : ~umbe~s. tunes. able takes that value. the random variable X is de. if the model is random variables. k. for i = I. by summmg the squares the chance that the random vari. One can . expected1)2/expected1 + (obwould balance.: 2. we value : would expect the n trials to rethe expected value of a random ~ s~t in o.II existentUdstll:tement I e:cpectedi =nxpi 57 *~~==~~~ ~ pected value of a constant a • existential statement a conditional that uses the word . For a discrete : tlStl~ summarises the discreprandom variable (one that has I anCles between the expected a countable number of possible : number of times each outcome values) the expected value is the ~ occurs (assuming that the weighted average of its possible ~ model is trU~) and the observed ~alues. The chisquared sta.I0f tIm~S an outcome of type i : ments.. gIve or take a bit. • .: observed i denote the number ' dependent repeated experi.~ well by the chisquared curve MsthlJllllilties============ II . + (observed k expected the expected value is a linear t ~)2/expectedk. As the sample operation: if X and Y are two : SIZe n mcreases.. Taking ~ + . The expected value of t occurs m the n trials. number of tImes. expected ~ if the model be correct. Let mg average of its values in in. where the weight as.each outcome sIgned to each possible value is : occurs. if it were cut out : served2 expected2)2/expected2 of a uniform material. the expected ~ c. a size change where k is greater t • expected i = nxpi than I • expectation. times a random variable X is the ~ constant times the expected 'same' : value of X (E(axX ) = aX • expansion I E(X».OITecr. and the ex.~ of t~e discrepancies. over all the categorandom variable as the point at ~ nes: chisquared = (observed1 which its probability histogram ..utcome i about nxpi ~ariable is the longterm limit.
I tions. closer to the area under the ChiSquared curve over the same range.~ grees of freedom (d.~ cess of RERANDOMI· SATION. _ experiment I what distinguishes an experi.exploration . in examining crop yield versus quantity of fertilizer applied.I ment from an observational I study is that in an experiment. and a protractor. The term is also used I to refer to the information necessary to describe the interre.f. In experiments.experimental design I this term overtly refers to the planning of a pJrocess of data I collection.indirect measurement clinometer a tool for measuring angle of elevation or depression. . This page illustrates the I sampling distribution of the chisquare statistic. .explicit form a formula for any term of a sequence given the number of the term. identification of repeated measures and replica. For example. the one that is observed is the dependent variable.I lationships within a set of data. _ expoential function to base a the function f(x)=ax• . I Such a description involves considerations such as number of ~ cases.58 txperiment I ~tment /I with (#categories 1) = k 1 de. the experimenter decides who I receives the treatment. sampling methods. and the crop yield would be the dependent variable. the explanatory variable is the one that is manipulated.I tification of variables and their I scaletypes. . iden. the explanatory or independent variable is the one that is supposed to "explain" the other. These considerations are I essential to guide the choice of : TEST STATISTIC and the pro. _ explanatory variable in regression. a plumb line. . x is called the base and y is called the exponent. consisting of an edge to sight along. sense that the chance that the : chisquared statistic is in any ~ given range grows closer and.). in the .exponent in the expression xl'. the quantity of fertilizer would be the explanatory or independent variable.
. not an equation. • face . the factorial operator is applione point. A factor model can be • exterior of an angle ~ characterised in terms of path the nonconvex set formed by an . : (of a polyhedron) one of the ~ polygons and its interior form~ ing the surface of a polyhedron. I : • factor (verb) • exterior angle I to fmd the factors of a number. . an angle that : technique for describing a set of forms a linear pair with one of ~ measured variables in terms of the interior angles of a polygon. Future Value as an Exponential FlIDction Amction 59 ~ • extrapolate . (of a polygon). FV : PV(l + If : PVoi"'J+'" = ed" y=FV a=PV b=lt(1+1l. ~ quantity. factor analysis is a multivariate 2. mula or a reasonable estimate.. tions. ~ factors. . with neither circle : cable to a nonnegative integer inside the other. I • factor (noun) • exradius ~ an exact divisor of a number. • externally tangent ~ • factorial (circles) intersecting at exactly . It is notated as the . diagrams to show relations beangle that measures less than : tween measured variables and 180 degrees. to use given information to pre: dict values beyond the set of ~ given values using either a for..·1 . 1. an angle that forms a linear pair with one of ~ • factor analysis the interior angles of a triangle. 1= .II '"Punmtislfunaiun Ifoetorial • exponential function the function f(x) =cX. an exradius of a triangle is the : This 7 is a factor of 28. • expression I • face angle combination of numerals or : the plane angle formed by adnumerals and variables that in~ jacent edges of a polygonal dicate a ftnite number of opera~ angle in space. (of a triangle).. radius of an escribed circle. a set of causal or underlying : variables.. y=ed" Relatillg to OUr c.
h u~ mcreas.Rs. but you must ante up Rs. .e v~ry ~ rapt y Wlty 1mcre~~ m . the false discovery rate is zero. .2 if a fair coin lands heads..al val ...Rs. I through preshape space. suppose I offer to pay you Rs.. fol~ lowed by a centered and scaled I • . bers of combinations.l to play.I + Rs.l + Rs.. sixths. at the origin and scaled to unit : centroid size) that differ only by ~ a rotation. For instance : 3! .~ ing value is the product of the . the false discovery rate is the fraction of rejected null hypotheses that are rejected erro foirbet IfibCl' II neously (the nwnber of Type I errors divided by the nwnber of rejected null hypotheses). increasing integer values from: I up to the value of the argu. By convention O! : is taken as producing the value ~ 1 f: . thirds. For example. This is a fair betin the long run.. fifths. nominators not greater than a : given integer.2x50% = Rs. if you ma~e this bet over and over agam. • fiber : the set of preshapes (configu~ rations that have been centered .2xP(heads) = .60 postfixed symbol 'I'. The result. . ~ • Fermat number I a nwnber of the form (22n+ I) : ~ • Ferma~s sp~al I a parabolic Sptral. you would expect to break even. is lx2x3 = 6. • false discovery rate in testing a collection ofhypotheses. Your expected payoff is .__ :1: fr cti" J.O. fourths. ~~on.OxP(tails) + Rs. argwthume~t va ue.raP l : gro IS represente 10 the ~ similarly rapid growth in num.~ ment quantity. tower or pyramid of power or importance farey sequence ~ the sequence obtained by ar: ranging in nwnerical order all ~ the proper fractions having de. with the convention that if no hypothesis is rejected. after accounting for the cost of the bet.CUlllllar a ons commonly used fractions such as halves. • family tree hierarchy. It is the path. ~ . eighths and tenths. ~ • fair bet ~ a fair bet is one for which the expected value of the payoff is zero. t diS .
.. 2. tIIe . associative for addition and multiplication.. 1. upon a mechanical system. As : applied in morphometrics...II Fibonacci numb". i... inverse for addition and multiplication. 1.'=======". :ill component particles) that are : expected to result from a speci~ fied pattern of stresses (forces) . FESA can be compared : with the thinplate spline.... : e.. . distributive for multiplication over addition. 3." fi~ nite element analysis is a comI putational system for con~ tinuum mechanics that esti: mates the deformation (fully ~ detailed changes of position of . Ifinite element ~'t!1=uurJ. commutative for addition and multiplication.. the 2por 3pspace of figures. ... • field properties closure for addition and multiplication...=:y... • Fibonacci number a member of the sequence 0. vectors. I . the landmarks. identity for addition and multiplication.. ~ • figure . _ mnabem 1ft the pd8m.""'SU.. a representation of an object by : the coordinates of a specified set ~ of points. • figurate numbers polygonal numbers ~ • finite element scaling I analysis : without the word "scaling.... coordinates under an entirely : different set of assumptions. where each number is the sum of the previous two numbers. ~ • figure space ... divided.6=1 configuration under all possible rotations. ""'wm .. T. ~ FESA solves the inverse prob.. the original coordinate data I . 5. which ~ interpolates a set of landmark .. lem of estimating the strains : representing the hypothetical ~ forces that deformed one speciI men into another. These results : are a function of the "finite ele~ ments" into which the space I between the landmarks is sub...
Sampling with replace. each such table will have a number of possible randomisations which may be a very large number. the smaller ~ the SE.. The SE for sampling . or the number of randomisations for each 2x2 table with the given row and II =======MJJthemlJnes . should be zero. there is no difference I between sampling with and without replacement.62 ~~~~~~~~* Jinitegroup I Fisher test II • finite group ~ exactly once). placement (every member of : the population is in the sample ~ • Fisher test named after the statistician RA Fisher. ~ rection gives (the numerator • finite population correc. Iation correction factor ((N n)/ : (N 1»'12. there is no variability in the re. including : the count in anyone of the 4 cells. pling without replacement is : called the ftnite population cor.. Such a test is useful to examine difference in rate between two binomial outcomes. the SE of sample sums ~ and sample means depends on I the fraction of the population: that is in the sample: the ~ greater the fraction. This is indeed : what the finite population corber of elements. without replacement is smaller : than the SE for sampling with ~ replacement by the fmite popu.. vanishes). This is an exact test to examine whether the pattern of counts in a 2x2 cross classification departs from expectations based upon the marginal totals for the rows and columns.. The randomisation set will thus consist of a number of tables with different respective patterns of counts.. ment. as in a simple random : sample. Note that for sample ~ size n=l..~ sult of sampling without re. For this test there are several reasonable test statistics. chisquared. tion when sampling without replace. If the sample size is the entire population of N units. and the SE a group containing a finite num. the fmite population correction is then I unity.~ rection. The randomisation set consists of those reassignments of the units which produce tables with the same row and column totals as the outcome. ment is like sampling from an : infinitely large population. The ~ adjustment to the SE for sam.
these are equiva. the tickets in the sizes are small.flexibly 1. and let p2 be the . _ Fisher's exact test . ones in the first sample has of zeros and ones.. Let . but would the appearance of computers. ~ are labelled with ones. Consider two populations .: samples would be expected to pergeometric distribution. If the null hy.~ number of tickets labeled with ages.I II Fisher'sl'&ll&ttest Iflexibly 63 column totals.. the two lent test statistics. This test has historically been ~ Conditional on G and the two regarded as superior to the use ~ sample sizes. placement from a collection of small 2x2 tables defined in : N = ni + n2 units of which G terms of row and column totals. under the null hypothesis. We would like ~ to set the ranges of observed to test the null hypothesis that : values' of the number of ones in pI = p2 on the basis of a simple ~ the first sample for which we random sample from each.~ pothcsis be true. of the sample from population : . sampl~s arc like one larger tion for the Fisher test is rela. tion of ones between the two ence to the algebra of the hy. exact test uses this distribution proportion of ones in the second population. hypothesis. Thus. Statistical tables : first sample are like a random have been published for the ~ sample of size ni without reFisher test for a number of . where students should be able G be the total number of ones : to mentally manipulate numin both samples. and nl. The calcula. making refer. Let pI be the ' (conditional on G) an hyper of ones in the frrst : geometric distribution with paProportion ~ rameters N. the for the equality of two percent. under the null of chisquared where sample . Let ni be the size ~ eSlS. would reject the null hypothpopulation. : have some chance variability. The allocacomputationally. sample from population 2. and let n2 be the size of the ~ usually applied to computation. sizes of the samples.: sample frum a single population tively undemanding ~ of zeros and ones. G. Fisher's population.~ bers and components of num . and ~ be proportional to the relative the test was widely used before .
64
bers to create a solution to a ~ • foot of altitude problem. i the intersection of an altitude of • floor function a triangle with the base to which I it is drawn. the floor function of x is the greatest integer in x, i.e. the ~ • foot of line largest integer less than or i the point of intersection of a line equal to x. : with a line or plane. • flowchart I a concept map that shows a: stepbystep process. Boxes I represent the steps, and arrows connect the boxes to order the : I process. • flowchart proof a logical argument presented in the form of a flowchart. I • focal chord a chord of a conic that 'passes through a focus. • focal radius a line segment from the focus • footballshaped scatterplot in a footballshaped scanerplot, most of the points lie within a tilted oval, shaped moreorIess like a football. A footballshaped scatterplot is one in which the data are homoscedastically scattered about a straight line. in morphometries, we represent the form of an object by a point in a space of form variables, which are measurements of a geometric object that are tIDchanged by translations and rotations. If you allow for reflections, forms stand for all the figures that have all the same interlandmark distances. A form is usually represented by one of its figures at some specified location and in some specified orientation. When represented in this wa~~ location and
i • form
c
of an ellipse to a point on the perimeter of the ellipse.
I
II = = = = = = = = = = M a t h _ t i & s
II form space Ifranu. mmpli"9franu *========6=5
orientation are said to have been "removed." _ form space th~ space of figures with differences due to location and orientation removed. It is of 2p3 dimensions for twodimensional coordinate data and 3p6 dimensions for threedimensional coordinate data. _ formula a concise statement expressing the symbolic relationship between two or more qu;mtities.  Fourier analysis in morphometries, the decomposition of an outline into a weighted sum of sine and cosine functions. The chapter by Rohlf in the Blue Book provides an overview of this and other methods of analysing outline data.  fourier series a periodic function with period 2 pi. _ fractal a selfsimilar geometric figure. _ fractal dimension a measure of the complexity of a structure assuming a consistent pattern of selfsimilarity (structural complexity at ; smaller scales is mathematically : indistiguishable from that at ~ largerscales) over all scales con; sidered.
I
BON
~
Size
Offset Width
540
110
2
I
.
5 6
8
3
..
13 22 33
44 55 66 77
2
2
Height
..
5 6
3
~
C!!!D
I Dd.ulll
7
8 9
9
O!!D
~
I
10
.10 12
" 12
I
88
111
"
00
:  fraction ~ an expression of the form alb . .
~  frame, sampling frame
I
a sampling frame is a collec
~ tion of units from which a
: sample .will be drawn. Ideally, I the frame is identical to the ; population we want to learn : about; more typicaHy, the ~ frame is only a subset of the ; population of interest. The : difference between the frame ~ and the population can be a ; source of bias in sampling de. sign, if the parameter of I terest has a different value for I the frame than it does for the population. For example, one
rn
MRthmuJti&s======= II
66
~~~~~~~=*
frelJumcy I funetion rule
II
might desire to estimate the ~ (fraction or percentage) of obcurrent annual average in ; servations in different ranges, come of 1998 graduates of : called class intervals. the University of Delhi. I pro I pose to use the sample mean income of a sample of graduates drawn at random . To fa I cilitate taking the sample and I contacting the graduates to obtain income information I from them, I might draw I names at random from the list of 1998 graduates for whom the alumni association has an I accurate current address. The I population is the collection of : 1998 graduates; the frame is I those graduates who have cur  frustum rent addresses on file with the for a given solid figure, a realumni association. If there is I lated figure formed by two para tendency for graduates with I allel planes meeting the given higher incomes to have upto solid. In particular, for a cone date addresses on file with the I or pyramid, a frustum is deteralumni association, that would I mined by the plane of the base introduce a positive bias into and a plane parallel to the base. the annual average income es I NOTE: this word is frequently timated from the sample by I incorrectly misspelled as the sample mean. frustrum. _ function rule _ frequency the number of times a value I the set of operations that deoccurs in some time interval. I scribes the process that takes the independent variable and _ frequency table transforms it into the dependent a table listing the frequency I variable in a consistent way. (number) or relative frequency
1J
.  
It represent the · lattice points in takes into consideration the the plane.==~============= • fundamental region a region used in a tesselation • fundamental rule of counting if a sequence of experiments or trials TI. • geoboard tion for two row vectors x. Defined by the equa.synonym for Mahalanobis distance. onto their consensus configuration. i. "nuclear stalemate" during the Cold War. in nl. . respectively. • gaussian curve a normal curve. A . Tk could result.. resistantfit. . . .:~ mill IIx  eil Vd. It is used to study economic behaviour. and a flat board into which nails • Xj for two individuals. where S is the pxp I rectangular pattern. T3. The fitring may involve leastsquares. I I dj(~) = p ceo} .. These nails variancecovariance matrix.. I • generalised superimposition the superimposition .ofa set of configurations. differences in directions in which there is less variation within groups are given greater weight than are differences in directions in which there is more variation. I I I I I I variance and correlation of the variables in measuring distances between points.nk do not depend on which outcomes actually occurred.. for example. .. n2 n3. e. and to model conflict between nations.(%)=~ I I I I I • generalised distance I d. • game theory a field of study that bridges mathematics. or other algorithms and may be strictiy orthogonal or allow affine transformations. and the numbers nl. statistics..67 . T2. . the entire sequence of k experiments has nl x n2 x n3 x x nk possible outcomes.. economics. nk possible outcomes. · . and p I have been driven in a regular variables as: . n2 n3. and psychology.
to landmark point locations. geometric morphometrics is a to and including the first suc. and length joining two given points. I II =~=======... eter p. . On I the side of a square having the a sphere. size.I same area as a rectangle whose tween two points as measured length and width are given. the the word "geometric" is the number of times one must toss geometry of Kendall's shape a fair coin until the first time I space: the estimation of mean the coin lands heads has a geo. the probability of success I The "geometry" referred to by in each trial.is' takes k trials to the first success. For example. it is the distance be. __. ________________. • geodesy a oranch of mathematics dealing with the shape. the geometric distribution de." . .MsJthmultiu .usually (but not always) limited pends only on the single param. metric distribution is lip. along a great circle.sample variation of shape using eter p = 50%. The multivariate part of px(l p)klto the event that it geometric morphometrics .I • geometric morphometrics scribes the number of trials ·up . curving paths have lengths)... • geometric mean • geodesic distance the length of the shortest path I the nonnegative number whose between two points in a suitable .=68= ' =======~* =D=eodeS'ie='=I=Beomem&=='morp==~=='= " ~ The expected value of the geo• geodesic the arc on a surface of shortest .~_ O____ ~~ :! 0 OJ ____ . and curvature of the Earth. ' usually carried out in a linear . The geometric distribution de.~ I i • geometric distribution .. in independent trials with I multivariate statistical analysis the same probability of success.. I of Cartesian coordinate data. The geometric I the geometry of Procrustes disdistribution assigns probability tance. square is the product of two geometric space (one for which given nonnegative numbers.collection of approaches for the cess.shapes and the description of metric distribution with param. its SE is (lp)V2/p..
geometric series I a series in which the ratio of : each term to the preceding I term is a given constant. . . . it is the class of morphometric methods that preserve complete information about the relative spatial arrangements of the data throughout an analysis. the branch of mathematics that : deals with the nature of space ~ and the size.geometry . As such.geometric solid ~ the property of a geometric the bounding surface of a 3di~ figure that it coincides with its mensional portion of space. properties of figures as well as : the transformations that preI . tangent space to the nonEuclidean shape space in the vicinity of the mean shape.IIBeometrie probllhility 1 probllhility IB~~1 symmetry . . and other .gergonne pomt I in a triangle. and other results in the space of the original specimens. . sample variation. expressed as the ratio of the number of successful outcomes of the event to the number of possible outcomes. • .given : information assumed to be true ~ in a proof ~ . : serve these propernes. I : . parallel to the translation vec: tor. . an isometry that is a composi: tion of a translation (glide) and ~ a reflection over a line that is .glide reflection . these methods allow for the visualisation of group and individual differences. . . the lines from the : vertices to the points of contact ~ of the opposite sides with the I inscribed circle meet in a point ~ called the Gergonne point.glidereflectional symme: try .geometric probability 1 probability (of an event) Its likely outcome. More generally. 69 ~ . shape.geometric progression I a sequence in which the ratio of : each term to the preceding ~ term is a given constant.
. . number) whose ratio to each : other equals the ratio of their ~ sum to the larger number. . often ~e~ to determine how steep a hill IS . Other randomisation ~ests may reasonably be Judged by comparison with this form. . .gnomon magic square a 3 X 3 array in which the elements in each 2 X 2 corner have the same sum. This Involves exhaustive randomisation. in such a way that each edge receives a different label. _ golden spiral a spiral through vertices of nested golden rectangles . _ golden rectangle . then there are always I : 2 prime numbers whose sum ~ is n I _ golden ratio . the ratio of two numbers .grace~ g~aph a graph IS Said to be graceful if you can number the n vertices with the integers from 1 to n and ~en label each edge with the dIfference between the numbers at the vertices. I I ~ : I : ~ I : ~ I : ~ .graph a gr~ph is a set of points (called vertices) and a set of lines (cal~ed edges) joinging these vernces.70 =======~.goldback's conjecture I if n is an even number greater : than 2. a rectangle in which the ratio : of the length to the width is ~ the golden ratio.gold standard the gold standard is the form of test which ~s ~ost ~ait?ful t? the randOI~llsatl0n distnbunon. (larger number : smaller. _ grad (0" grade) 1/100th df a right angle .grade the tilt of a reallife object in relation to the horizontal. image under some glide reflection. for a gIven test statistic ~nd experimental design.
x and yare : (x*y) *z=x* (y*z) for all X. . of them exactly. a graph of averages is a plot of the average values of one variable (say y) for small ran~es of values of the. verse element y such that . . ag~st the value of ~e s. squares sometimes used ~ measure distance to • great circle a circle on a sphere with a diameter equal to that of the sphere.econd vanable (x) at the ffildpomts of the ranges . and a binary operation * such : that ~ x*y is a member of G whenever . growing pattern the greatest common divisor of ~ a pattern where the number a sequence of integers.. '. The shortest path connecting two points on the surface of a sphere lies along the great circle passing through the points. the greatest lower bound of a set of real numbers. and ~ z .greatest common factor . ~ _ greatest lower bound . .IloraPhofaveraoes Igrowingpattem _ graph of averages for bivariate data. other van able (say x).graph theory the mathematics of complicated networks . : _ grid I ' : a tesse IatlOn 0 f congruent . ==========7=1 ~ . Y. x*y=y*x=e • greatest common divisor : . is the largI est real number that is smaller I than each of the numbers in the I set. is the I of objects in the pattern inlargest integer that divides each ~ creases from term to term. . the greatest number that is a : factor of each of the given num~ bers.group ~ a mathematical system consist~ ing of elements from a set G . there is an identity element e : such that e*x=x*e=e for all x ~ each member x in G has an in.
height l.L n i =1 Xi I • hectare a unit of measurement in the metric system equal to 10.47 acres). 2. I • I harmonic mean the harmonic mean of two numbers a and b is 2abj(a + b). • Hankel matrix a matrix in which all the elements are the same along any diagonal that slopes from northeast to southwest. the length of an altitude. • harmonic analysis the study of the r.000 square meters (approximately 2. • halfline a ray. • halfplane the part of a plane that lies on one side of a given line. (of a prism). • growth pattern a set of values usually visualised by plotting points on a grid and fitting either a linear or an expOnential equation to the scattel'" plot. The parameteric equation for a helix is x=a cos t y=a sin t z=bt .epresentation of functions by means of linear operations on characteristic sets of functions.the length of an altirude. • harmonic division a line segment is divided harmonically by two points when it is divided externally and internally int he same ratio.72 • growth formula either a linear ·or exponential equation that describes the growth over time. I • I I • I I I I I helix the path followed by a point moving on the surface of a right circular cylinder that moves along the cylinder at a constant ratio as it moves around the cylinder. Hannoni c Mean I 1 n HM==n 1 1 n 1 ni I L =1 Xi L  i = 1 Xi HM I I 1 1 n 1 =.
class intervals." A scatterplot or residual plot shows . a polyhedron having 6 faces.heteroscedasticity I seen in a drawing "Mixed scatter. . because the total area I . The .hexagonal number a number of the form n(2nl) .==========~====~ _ hemisphere ~ .heptagon a polygon with 7 sides. Linear regression is I a histogram is a kind of plot that not usually a good idea ·if the ~ summarises how data are disdata are heteroscedastic. and their heights . . That is. ~ bases of the rectangles are the . ~ .hierarchy heteroscedasticity if the scatter : a chart that shows varying levin vertical slices through the ~ els of importance plot depends on where you take I h" : .hidden lines : broken lines used to signify integer area. ~ dron.hexahedron half of a sphere including a . _ heronian triangle a triangle with integer sides and .1/ hemisphere I ~mm 73 . servations in the corresponding : class intervals. . : The cube is a regular hexahegreat circle as its base. ~ lines that normally wouldn't be .hexagon a polygon with 6 sides.Istogram the slice. ~ portional to the fraction of ob. a sixsquare polyomino. a set of rectangles ("bins") sit: ting on the horizontal axis. the histogram is . .hexomino . the ~ height of a given rectangle is I the fraction of observations in : the corresponding class interval. ~ divided by the length of the corI responding class interval. Starting with a set of ~ class intervals. : tributed.hexagonal prism : are such that their areas are proa prism with a hexagonal base. A his~ togram does not need a verti: cal scale.
organs or tissues. Experiments and studies with historical controls tend to be more susceptible to confounding than ." Following D'Arcy Thompson. This is called I the density scale.use many factors that might affect I the outcome other than the treatment tend to change over time as well.homeomorphism a onetoone continuous transformation that preserves open and closed sets. in studying the possible effect of fluoridated water on childhood cancer. such as molecules. for example.those with contemporary controls. The horiwn.homology the notion of homology bridges the language of geometric morphometrics and the language of its biological or biomathematical applications. . can be "homologous.historical controls sometimes. I I I I I I I I I I . with I controls to highlight ranges of ~ values and read their areas. beca.' homology II =========* of the histogram must equal I 100%.74 historical controls .) . The units of the vertical I axis are percent per unit of the horiwntal axis. only the explicit entities of evolution or development. In theoretical biology. the endpoint convenI tion is important. such as points or . (In this example. morphometricians often apply the concept instead to discrete geometric structures. If any observations coincide with the endpoints of class I intervals. the a treatment group is compared with individuals from another epoch who did not receive the treatment. Those individuals who were children before fluoridation started would comprise an historical control group. the level of other potential carcinogens in the environment also could have changed. we might compare cancer rates in a community before and after fluorine was added to the water supply. This page contains a histogram tool.I tal axis of a histogram needs a scale.
. . : size n (i.: x anon plate spline) a "homology map" ~ • hypergeometric ~stribumeans that one intends to refer tion . partial warp : where you take the slice. ~ from a population of N ob• jects of which G are "good. . about which we ." A scatterplot or ~ g good objects in such a residual plot shows • sample is GCg x NGCng! homoscedasticity if the scatter .._ 1. tures as if they had a consistent : • hyperboloid ' biological or biomechanical ~ a geometric solid whose equa2 2 2 2 2 ." • homoscedasticity : The chance of getting exactly "Same scatter. G and logical explanations pertaining : n is the distribution of the to the regions between the ~ number of "good~ obje'ct~ in landmarks. • hyperbola for corresponding parts in dif. a simple random sample of have no data.• . b2 =1.. .. the term "homolo. an rnterpol ' (such as a thin. and. In ~ • horizontal line this context. every subset of size n has the erators associated with the : same chance of occurring) specified structure. mearung.. provided g < = n.. most multivariate analyses. 2/a2 + y2/b 2_z2/C . processes affecting such struc. by a further exten.g.e.2= . a line whose slope is zero gous" has no meaning other than that the same name is used . S"lar1 to d ec Iare • tion is x /a + y2/b _Z /C = 1 0[. To declare something ~ • hyperbolic spiral "homologous" is simply to as. cion with parameters.f. ferent species or developmental stages. 1m1 y.~ in vertical slices through the sion..\I htmumwrpmsm I hypergeometric d~=tion========7=5 curves. C. to its features as if they had ~ the hypergeometricdi~tribu something to do with valid bio. a random sample • homomorphism ~ without replacement in which a function that preserve the op.: a curve with equation x2 2_ /a y2/ •. g < = MIJ.• the curve whose equation in sert that we want to talk about : polar coordinates is r*theta=a.n u = = = = = = = =___ 11 . scores) that arise as part of ~ heteroscedasticity.N. plot does not depend much on scriptors (e. NCn. to the multivariate de.
The other two sides are called legs. and its standard error is ((Nn)j(NI»V2 x (n X GjN x (IGjN) ) V2 . That chance is called the significance level of the test or decision rule. hyperplllne I hypothesis tutino I • I II G. volume to a space of more than three dimensions. hypothesis testing statistical hypothesis testing is formalised as making a decision between rejecting or not rejecting a null hypothesis. I • hypersphere a generalisation of the idea of a sphere to a space of greater than three dimensions. I • I • hyperplane : a kI dimensional subspace of I a kdimensional space. and n g < = N G.: perplane is typically characterised by the vector to I which it is orthogonal. it is possible to develop many different decision rules (tests).me . (The probability is zero otherwise. ~ a generalisanon of the ldea of . Two types of errors can result from any decision rule (test): rejecting the null hypothesis when it is true (a Type I error). • hyperspace a space of more than three dimensions.) The expected value of the hypergeometric distribution is n x GjN. A hy. The chance of correctly rejecting the null hypothesis when a given alternative hypothesis is . one specifies ahead of time the chance of a Type I error one is willing to allow. one way of deciding which decision rule is best is to pick the one that has the smallest chance of a Type IT error when a given alternative hypothesis is true. Typically. For any hypothesis. For a given significance level. and failing to reject the null hypothesis when it is false (a Type II error). on the basis of a set of observations. I I I : ~ . I • hyperv~lu.76 ~~~~~~~~~. I hypotenuse the side opposite the right angle in a right triangle.
or if both P and q : are true. Usually refers to a regular icosahedron. ~ical imPlica.1 empotent . I the element x III some algebraJ.II icoSahedron I implies. : • implies. one of the Platonic solids . . . structure IS called Idempotent if ~ x*x=x. h . I MAtbmuJncs======= II . • iff I if and only if.d .c. if P is false. • icosahedron a polyhedron with 20 faces. . imaginary Simplify: ~ . 77 true is called the power of the ~ tion ( (p AND q) OR ((NOT test against that alternative. then(p iff q) is a proposition that is true when both p and q are true. al and'l=sqrt( Xl were x IS re 1). . .. If is logically equivalent to the proposition ( (p IMPLIES q) AND (q IMPLIES p) ) and to the proposi I ~ logical implication is an opera I tion on two logical proposi~ tions. J( I)(.(p IMPLIES q) is ~ a logical proposition that is true . . I munber x+iy where x and y are : real is y. • identitiy transformation a size change where k equals 1 a). .) ' " I • identity reflection : • imaginary part a reflection where the preimage ~ the imaginary part of a complex and the image are the same. logical implication • iff. p) AND (NOT q» ). lent to the proposition ((NOT : p) ORq)./=4 . The proposition (p ~ IMPLIES q) is logicallyequiva. • identify to choose from a set or to name cases in which the desired resuIt is present or true. • imaginary number a complex number of the form . If p and q are two logical : propositions. if and only if if p and q are two logical propositions. • image : the reflection of the preimage I : • imaginary axis ~ the yaxis of an Argand diagram. and when both p and q are false. .
I • I I I I I independent and identically distributed a collection of two or more random variables {Xl.e. and are independent . the independent • included angle ~ variable is the one that is supan angle formed between two I posed to explain the other. a mirror) . independence two events A and B are (statisti~ally) independent if the chance that they both happen simultaneously is the product of tl:ie chances that each occurs individually. . term is a synonym for "explana• inclusive OR tory variable. This is essentially equivalent to saying that learn II = = = = = = = M J J t__ri&s ... a light ray) and I horizontal axis. independent variable in regression. Independent in the surface it rebounds against I this context does not mea. • incenter the point of concurrency of a triangle's three angle bisectors. and/ I gresses the "dependent variable" on the "independent varior able.. the (a cushion." Usually. i... if P(AB) = P(A)P(B). irukperulm&e II • improper subset a subset that includes the entire parent set.78==========im~7 subset I irulepnulmt... . X2. . The independent varipath of an approaching object able is usually plotted on the (a billiard ball. same thing as statistically independent . I • I I • independent. or both. . • incircle the circle inscribed in a given figure. one reone or the other. the given sides of a triangle.." There is not alwavs a • incoming angle I clear choice of the independent the angle formed between the variable. } is independent and identically distributed if the variables have the same probability distribution.
us I ~ ~ MlJthenuJrics======= II . Two random variables X and Yare independent if all events they determine are independent. • mdirect measure~en. and ~ zero if A does not occur.icaturrandmn variable I indirta. . and the standard error of I the indicator of A is (P(A) x (1: P(A)) V2.e. expected value of the indicator : of A is the probability of A. ~ .. every event determined by that subset of the variables is independent of every event determined by the variables in the complement of the subset. C.II iruJ. The product of the . is the random variable that : equals unity if A occurs.that IS lffipOS~ sible o~ lffipraCtlcal t? be me~ I sured directly or phYSICallY..t : a meas~ement . . For example. indicators of a collection of : events is the indicator of the ~ intersection of the events (the . The sum lA + IB + ~ 1C + .. C. . often written . . B.. ~ . b. and d.s"=rem&n==t========= 79 ing that one event occurs does not give any information about whether the other event occurred too: the condi tional probability of A given B is the same as the unconditional probability of A. lA. the three random variables X.. for example~ i~ the event {a < X < = b} IS mdependent of the event {c < Y < = d} for all choices of a. ~ P(A).. c. i. B. Y. collection of events {A. P(AIB) = P(A). } counts how many of t~e : events {A. } occur m ~ a given trial. of the indicators of a . product equals one if and only : if all of indicators equal one). and Z are independent if every event determined by X is independent of every event determined by Y an~ ~very event determined by X IS mdependent of every event determined by Y and Z and every event determined by Y is independent of every event determined by X and Z and every event deter~ mined by Z is independent of every event determined by X and Y : • indicator random variable ~ the indicator [random variable] ~ of the event A. ~ The maximum of the indicators . A collection of more than two random variables is independent if for every proper subset of the variables. The . of a collection of events is the : indicator of the union of the ~ events (the maximum equals lone if any of the indicators : equals one).
infinitesimal a variable that approaches 0 as a limit. I scribed in polygons or spheres inscribed in polyhedrons. I ally referring to circles in• infinite becoming large beyond bound. an indirect measurement.80 indirect proof I inst:ribetl1l1llJk " I • ~~~~=~~~=. of an angle starts from . I • indirect proof a proof that begins by assum. The I figure outside is circumscribed around the inscribed figure. . and maleI ing conjectures about • injection a onetoone mapping. curve is a point where the either the assumption or a pre. recognizing patterns. • inscribed angle an angle formed by two chords I of the circle with a common endpoint (the vertex of the angle). at which the tangent is changviously proved theorem. ally calculated using a formula or a known relationship. • initial side • inductive reasoning I the side that the measurement the process of observing data.curve has a stationary tangent..• inflection ing the conclusion is not true ~ a point of inflection of a plane and leads to a contradiction of . generalisations. I ing from rotating in one di• indirect technique the method used to determine I rection to rotating in the oppostie direction. Usuanother. • inscribed • inequality the statement that one quantity I (in a polygon or polyhedron) is less than (or greater than) Intersecting each side or face of a figure exactly once.
3 is the coefficient. or reasonable estimation. • interior of an angle the convex set formed by an • integer I angle that measures less than any whole number or its oppoSite. which there is no measurement of y is called interpolation. I • I internally tangent (circles) Intersecting at exactly one point. 3x. • integral coefficient in the expression.II instance ofa sentence I intersecting p:es===========""8. if I the value of x is within the range of the measured values of x. radius.1 • instance of a sentence ~ circle IS less than that of the a situation where the statement . to impute a value ofy corre~ sponding to some value of x at .thematics=================== II .. I • I interpolate determine a value within a set of given values using a formula. 180 degrees. is true . Angle A = Angle B interpolation given a set of bivariate data (x. I • I • interior angles angles between two lines cut by a transversal. rule. intersecting planes planes that share a line I I • MR. Integral coefficients are coefficients that are integers. imI puting a corresponding value of y is called extrapolation. with one circle inside the other. • interior of a circle the set of points whose distance from the center of the I • interquartile range the interquartile range of a list of nwnbers is the upper quartile minus the lower quartile. • intercepted arc an arc of a circle whose endpoints are marked by the sides of a central angle or an lllscribed angle. If the value of x is outside the I range of measured values. y)..
This im. I change under that transformation.iant an invariant."".tersec==~oftwosetSAllndB I i~verseoperation I II I I I fully be added or subtracted and that the mean is a representative measure of central tendency. In the morphometrics of elements contained in every one I triangles. c B. the san1e substantive meaning! I • inverse operation significance irrespective of the examples of inverse operations common level of the two val.. the invariants of a of the sets. and ratio of distances in the intersection of two or more I the same direction is an invarisets is the set of elements that ant under affine transformaall the sets have in common. ues being compared.I are addition and subtraction. {$} A = {$} AF= A rotation of one's coordinate sys• intersection I tern. I II ===================MJJthemllti&s . changes some inessential aspect anddistribttive of a measurement. lengths or weights . if not p.. the I tions.a form of conditional. • interval scale a characteristic of data such that • inverse the difference between two val.B or A. generally speaking." C. Hence invariant under translation or AA =. associative. For inAB=BA (AB)C=A. Such data are common in the domain of physical sciences or engineering e. Euclidean distance is an We note that if A. thenAB =A. is a quantity that is unchanged (even though its formula may I have changed) when one A.I multiplication and division. The intersection of : particular transformation are I the events A and B is written the shape variables that do not '1\ and B" and '1\B.alive.g.(BC) A(B+C)=AB+AC I stance.'1.2========'"". explies that scores may meaning• intersection of two sets A andB the set of elements which are in both A and B.f.'n"". • invar. ues measured on the scale has I then not q.8"".B The above operation is cOIIum. union..
or I tangent (of a number) The acute angle '. shapes that may contain curved portions rather than straightline segments . Standards Organisation.II in17eNe sine. I specifications are arranged gen• irrational number erally to concur with those of decimal number that never ANSI.ISO whose sine. The isometric drawing of a cube ~. cevians that are symmetric with : respect to the angle bisector. based in Geneva. the graph is a hyperbola with I for a number of standard programming languages. I shows all the edges equal. cosine. or tangent I isome. The axes as asymptotes.. but each square face is represented as a 60°120°60°120° rhomI bus. or tangent I acronym for the International is the given number. combinations of shapes. This • inverse variation I body publishes specifications a variation of the type xy = k. cosine. •• I • isometry : an isometry is a transformation of a geometric space that leaves Mllthemntics==================== II . I ends. cosine. Switzerland. never repeats (Ex: pi) : • isogonal conjugate • irrational number ~ isogonal lines of a triangle are a number that is not rational. tracting a root and raising to a power. gates if the corresponding lines to the vertices are isogonal. ments • isometric drawing • irregular shapes I a type of drawing that shows shapes that are not one of the named geometric shapes or I three faces of a threedimensional object in one view. I 83 • inverse sine. • irregular region region whose boundary is not the ~ Two points are isogonal conjuunion of circular arcs or seg. .
. board. tn. the isometnes (~or the I exactly two congruent sides. and • isosceles triangle reflections. If the distances are I a triangle with at least two conProcrustes distanc~s between I gruent sides. For. and the distances are distances between landmarks.84 isosceles tetrahedron I isotropic II =======* distances between points unchanged. landmarks ~n two : they are called the legs and the dimensions) are the rotatlons ~f I anoIe between them is called the Kendall's shape spat~ . these can be ~lsuahsed : the vertex angle IS called the as ordinary rotatIOns of I base.called the base angles." • isoscoles trapezoid • isosceles tetrahedron a trapezoid that has a pair of a tetrahedron in which each pair I equiangular base angles of opposite sides have the same I • isotomic conjugate length.• isotropic allel sides are congruent. two points on the side of a triangle are isotomic if t~ey ~re I equidistaru: from the. Isotropic errors have the same statistical distribu II =======MR.~ ve~ex angle. 1\vo pomts mSlde a triangle are isotomic conjuI oates if the correspon d. The nonvertex angles are Kendall's "spherical black. If a triangle has shapes. I ean translations. rru~po. The ~ide opposite angles.cs . I invariant with respect to direction. simplest case.mt of that side. • isosceles trapezoid a trapezoid whose two nonpar.th_ti. rotations. If the space is the Euclidean space of a picture or an organism. I cevians through these pomts meet the opposite sides in isotornic points .lllg b .
and (Xk ~ a unit of energy or ~ork. : . is in Ak) ). above the principal diagonal are : equal to 1..I 0 1 0 0 0 0 ability distribution deter.'S""S""hR""ifJ""esp=ac""e=~=====8~5 tion in all directions implying equal variance and zero correlation between the original variables (e. ments are O.. .II jointprobability distribution 'I Kenda. but all other eleI . each other.: where the left and righhand ability that each of two or ~ limits exist but are not equal to more random variables takes . axis coordinates). . 0 0 2 0 0 0 termined by the collection of I A:= 0 0 1 0 2 0 0 random variables: for any collection of sets of numbers 3 2 2 0 1 {AI. ~ . cal situation.... a discontinuity in a function a function that gives the prob.g.justify : give a logical explanation or in.Jordan curve ~ formal proof of a mathematia simple closed curve. the joint prob.Joule (X2 is in A2) and .Ale}. .I • mines P( (Xl is in AI) and: . their joint I 0 0 0 0 0 0 probability distribution gives 0 0 2 0 0 0 0 the probability of events de. a matrix whose diagonal ele: ments are all equal (and non~ zero) and whose elements . X2. I .. at a particular value. computation or : property. . . : underlying geometric ~ morphometrics.. ~ _ jump discontinuity _ joint probability function .. Kendall's MRthmuJri&s======= II .. . Xk are random variables.Jordan matrix _ joint probability distribution 0 0 0 1 0 0 0 if Xl. due to David Kendall. .Kendall's shape space ~ the fundamental geometric conI strUction.
. I I I I the crucial matrix for computing the thinplate spline interpolant between two landmark configurations.s"for the number of landmarks.' '_ WI. Each point in this shape space represents the shape of a configuration of points in some Euclidean space. Most multivariate methods of geometric morphometrics are linearisations of statistical analyses of distances and directions in this underlying space.000 meters . I I I I I I ~S . position. : ~ • knot . • kite a quadrilateral with exactly two distinct pairs of congruent consecutive sides.... ITDC I STfJt AU· DlJllHI7fIt AA4ItIIt ~ .1ONM "Ill" IMTZIW' .D . for historical II ======~MII""'tiu ..... ~~~ !'t~ . a ·curve in space formed by interlacing a piece of string and I then joining the ends together. .. . • kilometer a unit of length equal to 1. scatters of points correspond to scatters of entire landmark configurations.. In this entry. . • kinematics a branch of mechanics dealing with the motion of rigid bodies without reference to their masses or the forces acting on the bodies.. not merely scatters of single landmarks. In shape space..86 ~~~~~~~~=* kilonIeml I II shape space provides a complete geometric setting for analyses of Procrustes distances among arbitrary sets of landmarks. The angles between the pairs of noncongruent sides are I called the nonvertex angles. The angles between the pairs of congruent sides are called the vertex angles. k stand.. irrespective of size. and orientation. I • knight's tour I a knight's tour of a chessboard : is a sequence of moves by a I knight such that each square of ~ the board is visited exactly once.
The equation of the thinplate spline has coefficients Vlh.la lateral area • latera recta plural of lattice rectum. a vector of the xor ycoordinates of the landmarks in a target fo. : measured along the meridian I through that point. • lateral area the area of the lateral surface of a solid • lateral edge a segment whose endpoints are corresponding points of a cylindric solid's bases. ~ • latitude I . where h is. the surface not included in the : base(s) I : • latin square ~ an n X n array of numbers in I which only n numbers appear. or a face of a pyra~ mid that is not a base ~ • lateral surface ~ . or whose endpoints are the vertex of a conic solid and a vertex of its base the angular distance of a point ~ on the Earth from the equator.. Bending energy is the upper kbyk square of VI.m. followed by three D's (for two dimensional data. four D's for threedimensional data). : No number appears more than I . : • latus rectum I a chord of an ellipse passing through a focus and perpen ~ • lattice points . • lateral faces .themiJtUs======= II . the faces of the lateral surface : of a prism.87 *================~ reasons . points in the coordinate plane : with integer coordinates M. : once In any row or coIumn. The entries in the matrix L are wholly functions of the starting or reference form for the spline. I s I I • lateral edge (of a prism) The intersection of I two lateral faces.
for any triangle with angles of measure A. for every number e > O.. and c opposite C). .lab cos C. E > 0. then for every number E > 0. a. b~ and c (a opposite A. the difference between the number of successes and the expected number II =======MRthmMtiu . the percentage of successes is increasingly likely to be close to the chance of success as the number of trials lOncreases. . ~ o B. the chance that the percentage of successes differs from the probability p by more than a fixed positive amount. independent trials with the same probability P of success in each trial. (XI+X2)j2. This is equivalent to saying that the sequence of sample means Xl. if X!. . More precisely. I I I I I I "=1.) • law of averages the Law of Averages says that the average of independent observations of random variables that have the same probability distribution is increasinoIy likely to be close to the expec~ed value of the random variables as the number of observations grows. and C.. 0 • laW' of cosines the theorem that.Plural: latera recta. and sides of lengths . X3 .. . More precisely. are IIId ependent random variables with the same probability distribution. + Xn)j n E(X) I < E} converges to 100% as n grows. o I • I I I I I I I I law of large numbers the Law of Large Numbers says that in repeated. . Note that in contrast to the difference between the percentage of successes and the probability of success. X2 . I I law of contrapositive the type of valid reasoning that concludes the truth of a statement from the truth of its contrapositive .88 ~awofaverages I law ofla1lJenumhers II I • dicular to the major axis of the ellipse. (Xl+X2+X3)j3. and E(X) is their common expected value. b oppoSIte B. converges to zero as the number of trials n goes to infInity. c2 I = a2 + b2 . P{ I (Xl + X2 + . converges in probability to E(X)..
the button toggles between displaying the difference between the number of successes and the expected number of successes.II law ofsines I likelihood ratio test 89 *================= of successes. for any tri. MAthemtltics======= II . · : I. . and predicted sample values. The tool on this page illustrates the law of large numbers. e..• lemata • law of sines the theorem that. Many of the commonly • least upper bound the least upper bound of a set : used statistical tests are likeliof numbers is the smallest num.~ plural of lemma.~ hood ratio tests. The following tool illustrates the law of large numbers. to the likelihood when of every number in the set. and the analysis member of the set. true. angles with angles of measure lemma A~ B~ and C~ and sides of lengths ~ a theorem whose importance is a~ b~ and c (a opposite A~ b op. : Hotelling's T2. I0f vanance Ftest. I I . the ttest ber that is larger than every I for comparisons of means. sin sin A = sin B = sin C I • likelihood ratio test abc . ~ another. and the difference between the percentage of successes and the expected percentage of successes. n X p. sin I another theorem. • leg of a right triangle : a side of a right triangle that ~ include the 90 degree angle : • legs I (of an isosceles triangle).: primarily as part of a proof of posite B~ and c opposite C). ~ • leastsquares estimates . specified hypothesis is . tends to grow as n grows. the : two congruent sides of a non~ equilateral isosceles triangle. a test based on the ratio of the • least common multiple : likelihood (the probability or the least common multiple of a ~ density of the data given the set of integers is the smallest . parameter estimates that : minimise the sum of squared ~ differences between observed . parameters) under a general integer that is an exact multiple : model.g.
I between them that are on the tion. a : /1 line that passes closest to all of ~ .' them. / / II = = = = = = = M J J d l m u l r i u . The line segment connects the points. A straight arrangement of infinitely many points. I The measure of a line segment I is its length. y E I I I • line of symmetry the line of reflection of a figure having reflectional symmetry. I • line segment • line of reflection two points (the endpoints of the the line over which every point I line segment) and all the points of a figure is moved by a reflec.90 limit lline~t II =========* • limit the actual area of a region • line an undefined term in most deductive systems. line perpendicular to a plane a line perpendicular to every line in the plane that it intersects (or anyone of them) A r.~ v I IE • line of best fit I / .~1++7~ ··. Li __________. line containing them. ~ E ) i I • t. A line has no thickness and is therefore considered onedimensional.v L I\.___ j I given criterion.I given a collection of points. as measured by some .
. exconstant difference between I ample: By in a linear equation consecutive terms . We say that f is linear if for every number a and every value of x d fi hi h C() d C ) an y or w c l' x an 1(y are d efimed . : ~ .MiJt"'nu======= II .IS d efimed 1 l' d al £:() d ") £:( an equ s aX x. ~ • linear combination a sum of values each multiplied by some coefficient. two adjacent angles whose distinct sides lie on the same line. and that it makes sense to add things like x and y together.lilinesyrnmetry Ilineartenn of anequa. when applied to I • linear term of an equation consecutive whole numbers.I sociated with a proportional change in the other. . : ~ • linear operation suppose f is a function or operation that acts on things we shall denote generically by the lowercase Roman letters x and y. with the I f ' al I .I ciated if a change in one is as. <Angles 1 and 2 form a ) Linear Pair • linear function a function that. (') C( a x x ). • linear equation ax + By + C = 0 ~ ~. . same constant 0 propornon · h h h f : lty t roug out t e range · 0 I measurement. Suppose it makes sense to multiply x and y by numbers (which we denote by a). ~ • linear pair of angles . an (u x + y) is defined and equals f(x) + f(y) . A linear combination can be expressed as the inner product of two vectors. • linear pair 2 supplementary adjacent angles whose noncommon sides form a line.===========9=1 ~ • line symmetry a figure has line symmetry if there is a least one line that divides the figure into two parts that are mirror images of each other. : • linear association two variables are linearly asso. . one representing the data and the other a vector of coellicients. but no generates a sequence with a : exponent m an equauon. the term w~th a variabl~. The corre lanon: coefficient measures the degree I of linear association on a scale : of 1 to 1.
I intuitive way. that can be ear transformation is the con.". includI dent variable. one tor space (what is "five times a linear transformation takes I point"?).I tiplied by real numbers in an abIes that are all linear combi. takes them to relativewarp I • lines of sight scores. the most I a description of the values of a common kdimensional linear function or relation within a vector space is the set of all real I small interval of the indepenkdimensional vectors.one of the four main descripear vector space is a set of ~le. In plane don't form a linear vecgeometric morphometrics. with the • loading I the correlation or covariance of rows of A. coefficients supply formulas for the compuI tation of scores whereas loadI ings are used for the biological interpretation of the linear combination .added together and can be mulstruction of a new set of vari. usually bits of geometry in multivariate statistics.92===========* lineartransfOrmatUm IlocatUm II • linear transformation I ments. A linear transformation I lines from an eye to what it sees of a matrix A can be written in that show perspective 'and what the form y = Ax. x. another origin do form such a space. a lin. • linear vector space • local behaviour in morphometrics. More I • location generally. but informally. where y is the ~ size to draw it resulting linear combination of .. a lin. In general. ing all sums of these vectors and their scalar multiples. a column vector.I or whole functions. A x j I loading is not the same as a coI efficient.tion of a point II =======MlJthenuuies . but lines segments Procrustesfit coordinates to connecting all the points to the partial warp scores. a measured variable with a linear combination of variables. The points of a nations of the original set.
• locus the path of a moving point. For each of a set of values of the predictor variables. and ending with a final statement called a concluI I sion. • logical argument a set of premises followed by statements. eastsquares regreSSIOn method. the effect of aging on some measured variable. • longitudinal study a study in which individuals are followed over time. randomisation tests for this purpose have been developed by Mehta & Patel. for example. the set of all points in a plane satisfying some given condition or property. • logarithmic growth a set of values which are approximated by an equation of the form y = log b x. Fitting is via a weighted 1 . the outcomes are regarded as representing a binomial process.111OCRtion. measureofllongitudinaI7==========~9=3 • location. with the binomial parameter 'p' depending upon the value of the predictor variable. each of which relies on the premises or on previous statements. Longitudinal studies provide much more persuasive evidence about I I I I I I I I I I I I I I I Mathemlltics==================== II . and cornpared with themselves at different times. The modelling accounts for the logarithm of the odds ratio as a linear function of the predictor variable. • logarithmic notation use of the symbols "log" or "in" in context. • logistic regression this relates to an experimental design for predicting a binary categorical (yes/no) olltome on the basis of predictor variables measured on interval scales. to determine. • logic the study of the formal laws of reasonmg. • logarithm a logarithm of a number is the exponent to which a given base must be raised to produce the given number. measure of a measure of location is a way of summarising what a "typical" element of a list isit is a onenumber summary of a distribution. An argument is valid if the conclusion has been arrived at through deductive reasoning.
4 I 3 5 8 1 9 2 7 6 I • I • ltetromino a tetromino in the shape of the letter L. a curve that cuts all parallels under the same angle. If the successive squares of a tour on an n X n chessboard are numbered from 1 to n A 2. where each number is the sum of the previous two I I I I I magic tour if a chess piece visits each square of a chessboard in succession. 7.94 loJl1er bountl I I I mRBie tour II =========* the effect of aging than do crosssectional studies. or main diagonal is a constant (known as the magic sum). . • lowest common denorninator the smallest number that is exactly divisib.. II =======MAtlmnRtics . L} =1. 1. in order. I I • magic square a square array of n numbers such that sum of the n numbers in any row. the tour is called a magic tour if the resulting square is a magic square. • lower bound any number below which a function value may approach but not pass. 4. 3. . column. numbers. Lo =2. Ln =Ln_ } + Ln_r I • I I lune the portion of a sphere between two great semicircles having common endpoints (including the semicircles).loxodrome on a sphere.. • lucas number a member of the sequence 2. this is called a tour of the chessboard.1e by each denominator of a set of fractions.
: tance from the origin ~ • magnitude of a rotation the amount of rotation in de. Reference srs1ltml • malfatti circles three equal circles that are mutually tangent and each tangent to two sides of a given triangle. The test statistic is ~ a measure of . • MannWhitney test this is a test of difference in location for an experimental design involving two I samples with data mea: sured on an ordinal scale or ~ better. its dis. MsthmuJties======= II . built. ordinalscale data but is not : confined to this scaletype. This test : is generally recommended ~ for comparisons involving . ~ An equivalent formulation . of the test.. ~ the numbers in these three .<?:rdinal prece. and its image • major arc an arc whose endpoints form an angle over 180 degrees with the center of the circle. based upon : ranking the data and sum~ ming ranks within groups. . dence. : the pair is classified in one ~ of three ways according to . is the Wilcoxon test. The ~ randomisation set is the I same as that for the Pitman ~ permutation test. For each possible : pairing of an observation in ~ one group with an observa. grees : • magnitude of a translation the distance between any point ~ 95 *================ • manipulatives objects that can be arranged. zero or negative. tion in the alternate group. categories are tallied over : the randomisation set. and moved around by hand. • major axis the major axis of an ellipse is it's longest chord.II magnitude IMannWhitney test ~ • magnitude the value of a number. written the extra letter is used to distinguish it from a minor arc.~ . whether the difference is : positive.
or equation) representing a realworld situation. t h e m s r i e s . The margin of error of an esti. • mathematical model a mathematical q. For random variables: if a random variable X must be nonnegative. the related mathance. Then the maximum likelihood estimate of p is estimate p by the value q that makes P(observing x when the value of p is q) as large as possible. divided by a. an estimate of a parameter. graph.MANi~. I • mathematical notation unfortunately.I a model. symbols. correct use of labels. For example. the fraction of numbers in the list at least as large as any given constant a>O is no larger than the arithmetic mean of the list. ematical problem is posed and I solved.!!!!!!96~~~~~~~~==. suppose that the parameter is p. In mathematical modeling the real • maximum the largest of a set of values.~ and abbreviations in a mathI ematics contett. mate is typically one or two times the estimated standard I • matrix arrangement of pixels error of the estimate. and the mathematical . and that we observe data x.bject (such as a geometric fig~re. than the expected value of X. suppose we are trying to estimate the chance that a (possibly biased) coin II = = = = = = = M A . not everyone agrees what it should mean. I I I I I I I I • maximum likelihood estimate the maximum likelihood estimate of a parameter from data is the possible value of the parameter for which the chance of observing the data largest.OVA I mmeimumliltelihootlestimtlte II .mapping I solution is interpreted back into making a transformation the realworld situation as a • margin of error solution to the realworld proba measure of the uncertainty in I lem. That is. • Markov's inequality for lists: If a list contains no negative numbers.MANOVA ~ world situation is abstracted to see multivariate analysis of vari. the chance that X exceeds any given constant a>O is no larger. table. divided by a.
The distribution of the number of times the coin lands heads is binomial with parameters n (known) and p (unknown). • mean. ~ . the parameter t. : ~ i : ~ ~ . The maximum likelihood estimate of p would be the value of q that makes that chance largest. Thus the maximum likelihood estimate of the chance of heads from the number of heads in n independent tosses of the coin is the observed fraction of tosses in which the coin lands heads. if X is an estimator of . where the term I is in the antecedent. I. • mean average • mean squared error the mean squared error of an estimator of a parameter is the expected value of the square of the difference between the estimator and the parameter. Our data will be the number of times x the coin lands heads in n independent tosses of the coin. : • measure of an angle I the smallest amount of rotation ~ necessary to rotate from one : ray of the angle to the other. measure the amount of openness in an : angle ~ I I. the fraction of times the coin is observed to land heads in the n tosses. ~ divided by the number of numi bers.meaning : a version of a conditional that I : defmes a term. al versus mterv scale versus ra Mathematics================= II . The main distinctions are concerned with the meaningfulness of numerical comparisons of data (nominal scale versus ordinal scale . it turns out to be q = xjn. The chance of observing x heads in n trials if the chance of heads in a given trial is q is nex qx( lq)nx. See also average. In symbols. I : • measurement type this is a distinction regarding the relationship between a phenomenon being measured and the data as recorded. ~ usually measured in degrees. I • measure of an arc ~ the measure of minor arc or : major arc is the measure of its ~ central angle. then . arithmetic mean : the sum of a list of numbers.II mean I meamremmttype ~ 97 *==~~====~~== lands heads when it is tossed. We can find that value of q explicitly using calculus.
mersenne number a number of the form 2pl where p is a prime. : • median ~ 1. . "Middle value" of a list.98 . and whether the scale is bounded or unbounded. The median can be estimated from a histogram by fmding the smallest number such that the area under the histogram to the left of that number is 50%. median of ~ interlandmark distances. IT the ~t I has an odd number of entrIes. (of a triangle). • method of comparison the most basic and important method of determining whether a treatment has an effect: compare what happens to individuals who are treated (the treatment group) with what happens to individuals who are not treated (the control group). The . .I peated . whether the scale of the measurements (other than nominal scale measurements) should be regarded as essentially continuous or discrete.rs m the ~t ~ are no greater than It.at : least half the numbe. I • I 8 I • c I • ~edian size a SIZe measure based on the re. 2. Used· in resistantfit methods. smallest number suc~ that .w I I mangle rmethod ofcomptlrison II tio scale this is known as Stevens' typology). • medial triangle the triangle whose vertices are the midpoints of the sides of a given triangle. member of a set something is a member (or element) of a set if it is one of the things in the set. A I I I I I in the list after sorting the list into increasing order. : the median is the middle entry ~ I • mersenne prime a Mersenne number that is prime. IT the list has an even number of entries. the median is the smaller of the two middle numbers after sorting. a line segment connecting a vertex to the midpoint of the opposite side.
with the center of the circle. • minimalchange sequence : exploration of a randomisation ~ distribution is such a sequence I that the successive : randomisations differ is a ~ simple way.de s~gmen} ~ an angle less than 180 degrees conn~dctmg t e ml pomts 0 . .1\ metric space I minor lire 99 ~ • minimal path .I the smallest of a set of values. same distance from both end. • metric space a space and a distance function defined on every pair of points that meets the requirements of the definition of "metric" above. the path of shortest length. as : when fmding the shortest path ~ from one point to another by .A. the line seg. ~ .(of a trapewid).• mtmmum 1. 0 a t: ~ a . way of a fIxed line. and further promoted by G. : • midpoint (of a line segment) The point ~ on the line segment that is the . • midp proposed by H. . whatever the opponent might do (whatever strategy the opponent might choose). MathBm4tics======= II . ment connecting the midpoints I • minor arc.. : two Sl es • minimax strategy in game theory.O Lancaster. an arc w ose en pomts form .: points. ~ Ods I ° ° • mt egment .· h d· 2 (f nangl) lin . of the two nonparallel sides. Barnard. This is a tail defInition policy that the alpha value should be calculated as the sum o( the proportion of the tail for data strictly more extreme than the outcome. In the context of a I ranodmisation test this can ~ mean that the value of the test : statistic for a particular ~ randomisation may be calcu. lated by a simple adjustment to : the value for the preceding ~ randomisation . plus one half of the proportion of the distribution corresponding to the exact outcome value. This gives an unbiased estimate of alpha. a minimax strategy is one that minimises one's maximum loss.
if the list consists of the N elements xl. I . B(Xk).100 ~ minor IlI4ds I monic polynomi. • model to create.lll II a representation of a mathi ematical relationship or situ: ation. . The kth moment of a random variable X is the expected value of Xk.. drawings or symbols.. For histograms. A list can have more than one mode. I • modus tollens the type of valid reasoning that uses "if P then Q" and statement "not Q' to conclude that "not P" must be trUe. II =======MJJthem4tics . x2. • mixed variation variation that contains both direct and inverse variation. I • minor axis the minor axis of an ellipse is its smallest chord. I monic polynomial a polynomial in which the coefficient of the term of highest degree is 1. • modus ponens the type of valid reasoning that uses "if P then Q' and the statement P to conclude that Q must be true..mira a plastic device which is used to determine and complete symmetries by reflecting images and allowing the user to also see through the reflecting surface. I • I I I I I moment the kth moment of a list is the average value of the elements raised to the kth power. using concrete materials. the kth moment of the list is ( xlk' + x2k + xNk )fN. • mode for lists. . I : • modulo I the integers a and b are said to I be congruent modulo m if ab is divisible by m .xN. . the mode is a most common (frequent) value. that is. a mode is a relative maXImum ("bump").
• morphometrlcs function f decreases monotoni. and ~ characterised by what aspects of strictly monotonically decreas. . . ~ of the measurements once they named after the famous SIte of . are made. : the values of the test statistic to ~ generate an estimate of the • monohedral tiling . This procedure is are c.~ glossary emphasise configuracarlo test mvolves generaong a ~ tions of landmarks from whole random subset of the. An advantage I of the MonteCarlo test over the • monotone a sequence is monotone if its ~ bootstrap is that with succesterms are increasing or decreas. ~ in contrast to the bootstrap pro• monomial I cedure in that the sampling of an algebraic expression consist.• monochromatic triangle . implies thatf(x) > f(y). casinos. to measure. : out replacement. A m~nte. f ~ an effective pseudorandom Increases monotOnIcally (IS : generator.: sive resamplings it converges to ~ the gold standard form of the mg.: biological "form" they are coning if if x > y. ." and : "metron' " meaning "measureif x > y." Schools of monotonically increasing if x > I morphometrics are y. implies thatf(x) ~ cerned with. : for the MonteCarloprocedure a funct10n IS monotone if It only I is a source of random codes or ~ncreases or only de~reases:. and using colored the same. ~ randomisation set." cally (is monotonic decreasing) I meaning "shape. and what kinds of : biostatistical questions they ask • MonteCarlo test . ~ implies thatf(x) > = f(y). what they choose < f(y). sampled a triangle whose vertices are all . exact test. . form of the full randomisation a tessellation in which all tiles : distribution. . An effective necessity • mon~to~c funCtlO~ . without replacement. The methods of this gambling. from the Greek: "morph. implies thatf(x) < = I f(y).. A function f is strictly : ment.~ the randomisation set is withing of just one term. Msth_ties======= II .. A . monotonic increasing) if x > y.ongruent.
2.I an algebraic expression consisttions with each other and with ing of 2 or more terms. size. it is trying to estimate. the method of "allometry. k.. It is a summary mea. There • multinomial distribution are many others.( (#&')2) I multinomial tlistributiun II organs or organisms analysed ~ of the bias and SE of the estiby appropriately invariant bio. and . so pI + p2 + . sure of the accuracy of the es.come in category j. A third." measures sizes of separate organs and asks I • multinomial questions about their correla. each of the MSE measures how far which can result in an outthe estimator is off from what come in any of k categories. cause or effect with I • multimodal distribution position in Kendall's shape I a distribution with more than space) in order to answer bio. taxon.I more than one "bump. In partruth (bias). The n trials result in nl outcomes MSE can be written in terms I of type 1. n2 outcomes of type II =======MathimuJries .... and the variabil. mator: MSE(X) = (bias (X) )2 metric methods (covariances of : + (SE(X))2.each trial results in an outments..102 mse(x) =.I ticular. The number of outdeney of the estimator to comes of each type has a mulovershoot or undershoot the I tinomial distribution. the probability that the ity of the estimator (SE).lone mode. on the I Let pj be the probability that average in repeated experi.100%. + pk = timator. Another sort multimodal distribution has of morphometries studies tis. j = 1.I . measures the densities of points and curves.4 uses these patterns to answer I •• •• questions about the random processes that may be controlling the placement of cellular I 4 structures." sue sections. measures of total size. consider a sequence of n in• mse(x) = e( (Xt)2 ) I dependent trials. It combines any ten. The histogram of a logical questions. .
. ~ • multiple the integer b is a multiple of the integer a if there is an integer d such that b=da..e. .multiplication rule the chance that events A and : morphometric analysis. SomeB both occur (i.multivariate ~ morphometries .. .. . is the conditional I morphometric technique must : be multivariate in nature. I the prediction of two or more ~ dependent variables using two ~ MIIthernatics======= II . . and nk outcomes of ~ type k is n!/(nl! x n2! x . n.~ negative integers that sum to . : . times the un. _ multiple regression the prediction of a dependent variable by a linear combination of two or more independent variables using leastsquares methods for parameter estimation.nk are non. : X pknk. An analysis ofvari~ ance of two or more dependent . °a term historically used for the : application of standard multi~ variate techniques to measure. that event ~ what confusing now as any AB occurs).multivariate analysis of variance ~ MANOVA.II multiple I multi"'ariatemultipleregr~ 2. : . . . I 103 • multiplicity in hypothesis tests in hypothesis testing. variables considered simulta~ neously.•. if more than one hypothesis is tested. probability that A occurs given I that B occurs. . . . or more variables per indi~ vidual.multivariate multiple I • conditional probability that B : regression occurs.: Wlse. ment data for the purposes of . X nk!) x plnl X p2n2 X .multivariate data I a set of measurements of two . the actual significance level of the combined tests is not equal to the nominal significance level of the individual tests . th~ chance is zero other.: . _ multiple discriminant analysis discriminant analysis involving three or more a prioridefmed groups. ifnI.
. . original statement is false. and true if the .. a population of numbers (a list : of numbers) is said to have a I . • nadir the point on the celestial spehere in the direction downwards of the plumbline. .. the lines from the vertices to the points of contact of the opposite sides with the excircles to those sides meet in a point called the Nagel point. I • negation /' I (of a statement) A statement : that is false if the original state~ ment is true. " A double negation of a state• natural number anyone of the numbers 1. • nagel point in a triangle. • navigational system compass directions or bearings in a variety of formats.. Then the list is I nearly normally distributed if.. • multivanate regression the prediction of two or more dependent variables using one independent variable. ment is a negation of the negation of that statement. . I . the fraction of numbers in the I list that are between a and b is I approximately equal to the area under the normal curve beI tween (a JL)/SD and (a JL)/SD. I • negative binomial distribution consider a sequence of independent trials with the same . supI pose that the mean of the list is : JL and the standard deviation of ~ the list is SD. for every two numbers a < b.!!::!lO!!::!4=====!!::!m!!::!ul!!::!tm!!::!' I1II'!!::!Ut'!!::!te!!::!regremon: I negstiPe binomiRldistrihution !!::!' *". I histogram of its values in stan: dard units nearly follows a nor~ mal curve.3. or by pre: ceding the statement with the phrase "It is not the case that. • . II or more independent variabies.2. 4. More precisely. ~ • nearly normal distribution .5. The : negation can usually be made by ~ appropriately adding the word I not to the statement. nearly normal distribution if the .
. M/lthemtI..I ter event is p.. • negative number a number smaller than O..: mial distribution with param. The number of tri. r+2. there must have been r.I eters n and r.I eters n=k1 and p. als and the kth trial must re.~ 1 successes in the first k1 tri. the former is cite chance of r. ~ . has probability k1Cr1 x prl x (1p)kr. because there must be at least r trials ~ to have r successes. tive binomial distribution is I derived as follows: for the rth : success to occur on the kth ~ trial. r+ I. success has the negative Bino. • net a twodimensional pattern that you can cut out and fold to form a threedimensional figure. • rune pomt circle in a triangle.: p)kr.I ity of success in each trial. which.. and zero for k < r.: 1 successes in k 1 independent ~ trials with the same probabil. the circle that passes through the midpoints of ~e sides is called the nine point CIrcle.. . according to the Binomial distribution with param..tics=======:. If the random : variabJe N has the negative ~ binomial distribution with . .. for k = r.~ suIt in success The chance of .:' II ... parameters nand r. the circumcenter of the medial triangle is called the nine point center. . The nega. 105 probability p of success in ~ each trial. by assumption.. which gives the expression for P(N =k) above. then : I P(N=k) = k1Cr1 x pr x (1. als up to and including the rth . Because the trials are independent. we can find the I chance that both events occur I by multiplying their chances together.. The chance of the lat.II negllti:"enumbQ' I ninepointcircle . . • network a group of nodes and arcs • ngon a polygon with n sides • nine point center in a triangle..
The fine : detail of the comparison invol ves the tail definition I policy. such that the I outcome of a statistical test is I classified in terms of whether I the obtained value of alpha is : extreme as compared with I this criterion level. • nonagon a ninesided polygon. : The commonest conventional I values for the nominal alpha I criterion level are 0. • nominal alpha criterion I level I a publicly agreed value for typel error. if the situation is an observational study rather than an experiment).~1~06~~~~~~~~~no~M~~~wi~mdm8ni~~I~mJ • no causation without manipulation a slogan attributed to Paul Holland.01.05 and 0. other. the collection of languages in which an international treaty is published. ~ • nominal scale : I I I I this is a type of measurement scale with a limited number of possible outcomes which cannot be placed in any order represenring the intrinsic properties of the measurements. Examples : Female versus Male. If the conditions were not deliberately manipulated (for example. 0cId_ . • node a description of a point in a network where it is possible for two different segments to share the same endpoints II I ~ alpha criterion level. it is unwise to conclude that there is any causal relationship between the outcome and the conditions. wise not ('nonsignificant'). The outcome is classified as showing statistical significance ('significant') if the I outcome has low alpha as compared with the nominal • nomograph a graphical device used for computation which uses a straight edge and several scales of numbers. See post hoc ergo propter hoc.
change by a constant value as the value of the independent I • nonoverlapping regions variable changes by a constant ~ regions that don't share interior : points value. eral of these tests were : explictly concerned with ordi• noneuclidean geometry ~ nalscale data for which modsolid geometry I eling based upon continuous • nonincluded side ~ functions is clearly inapprothe side of a triangle that is not : priate. Sevto the given line. the relationship between two 2. the set lie entlreIy In the set. mostly over the segments connectIng POInts of. . With the . that is. second is not simply proporset of data or table of values : tional to the change in the: first. in which the amount of the ~ independent of the value of the dependent variable does not I first variable. through . . spec! !C 0 Jectlve synonym: concave. These tests are implicincluded by 2 given angles ~ itly rerandomisation tests.nonary . . perlOd 19301960 . a change in the other that is associated with 9 : depends on the value of the • nonconstant rate of ~ first. .. ing assumptions about sam• noneuclidean geometries : pIing from populations with hyperbolic geometry ~ data supposedly conforming a geometry in which. I °fi b' . 0 f bypass. : variables is nonlinear if a ~ change in one is associated with . . I • : • nonparametric test • non~on~ex ~et hi h all I a number of statistical tests a set 0 POInts In ~ c ~ot ~ were devised. 0 MIIthenumes======~ II . if the change in the change . there are : tistical distributions such as infinitely many lines parallel ~ the normal distribution.II nf11'Ull9onal number 11I01IIPfJ//rametric:==========~1~O~7 ~ • nonlinear association • nonagonal number a number of the form n(7n5)/ . to theoretically modeled staa point not on a line.
. Many probability distributions can be approximated by I I I I I ~ • normal approximation I : ~ I ~ : ~ II =======MRthemtlBes . • normal perpendicular the normal approximation to data is to approximate areas under the histogram of data. • nonvertex angles (of a kite) The two angles between consecutive noncongruent sides of a kite. the survey may suffer from nonresponse bias. everyone who is mailed a questionnaire. • nonresponse rate the fraction of nonresponders in a survey: the number of nonresponders divided by the number of people invited to participate (the number sent questionnaires. Nonresponse bias makes the result of a survey differ systematically from the truth. many hours people work is likely to miss people who are working late. congruent. in ways that are responds. When that happens. is the a telephone survey of how nonresponse. it is rare that everya shape that does not have all lone who is "invited" to particisides congruent and all angles I pate (everyone whose phone . The difference berelated to the effect one is try. For example. transformed into standard units. those who respond may differ from those I stop on the street . by the corresponding areas under the normal curve. ) in fact who do not. I sought. number is called. the number of interview attempts.) If the nonresponse rate is appre I • nonrigid transformation a transformation that does not preserve the size and shape of the original figure. .108 nonpersp:edrtllwing I1UJrmtJl4ppyoximation I II I • nonresponse • nonregular shape in surveys. and that obtained.I tween the "invited" sample ing to measure. the survey suffer from large nonresponse bias. etc. and are therefore not at home to answer the phone. ev• nonresponse bias I eryone an interviewer tries to in a survey. • nonperspective drawing a threedimensional drawing that doesn't use perspective I ciable.
. To use the normal ap. : cable for continuous interval Msth_tUs========== II . The area normal approXI.. al f ul b I pOSItiVe lor every v ue 0 x.5 and 17. I c. ~ iar "bell curve:. The mathematical the binomial probability his~ expression for the normal togram for n = 50 and p = : curve is y = (2xpi)V2Ex2/2. rallogarithm (2.~ not most) histograms con.24 =. (n x p x (lp)) 1/2 = . . See .Th al .~ ity histogram. in the ~ The tool on this page illussense that the area under the .77. : .. n X p = . c. pected value (for the Bino.. mal curve.II nomudcurve I nomuddistributUm *=========1~O~9 a normal distribution.24 = 1. 0 2 15)/3 .71828 . mla ran om vana e. verte d Into stan d ar d'units. approximation gets worse curve. by subtracting the ex. mation is the area under the .and 15 fior these vaIues 0 f nan· d .I d . t h at area IS I 73. trates the normal approximaprobability histogram is close : tion to the binomial probabilto the area under a corre. Note that the sponding part of the normal .di th ) an p Vi ng e res t .697 and (17." illustrated on For example.around t h e pomt x .: when p gets close to 0 or 1. slightly smaller than : • normal distribution the corresponding area under I the normal distribution is a the binomial histogram. ). 5 normaI curve b etween (9 . nc. .24 for these values of n an~ : fined.: the normal curve is the familviding by the standard error. cumference of a circle to its proximation. the area under I this page. 30% between 9. To fmd the correspond. and the SD of nOffilal. ~ and that the approximation the range must be converted ~ improves as n increases. we transform : d·lameter. bl : e norm curve IS symmetI .5 is ~ where pi is the ratio of the cir74. 11 th 15)/3.. is also unit~ Many (but p).. (3 the endpoints to standar d ~ and E is the base of the natuunits.5 I approXimate y 10 ow e nor. . 14159265 .5%. the normal curve suitably de3.0 . ) . . by subtractI • normal curve ing the expected value and di. to standard units.~ curve is unity.~%. .y : The area under the normal the ~tandard error (for a BI. d di . ing part of the normal curve. theoretical distribution applialso the continuity correction.
these samplmg distributions are the chara~ten~t~c t~ols of paramet. It is ~ the unexpectedness of the de~ gree of departure of the ob. morphometrics. F.. It is related mathematically to the binomial and chisquare distributions and to several named sampling distributions (including Student's t. it is usual to I propose a null hypothesis that : the pattern does not exist. . by the measure alpha. the par~eters : for translation and rotation are ~ usually nuisance parameters. which is examined . in order to test whether a sup: posed interesting pattern exists ~ in a set of data. • normalise to normalise a geo~etric object is to transform It so th~t some function of its coordInates or other parameters has a prespecified value. That is. In . vectors are of~en normalised by transformation into unit vectors.the simplest model under conI sideration. Pearson's r). relative to the pat~ tern expected under the null ~ hypothesis. th . • nth term . : • null model ~ . ReferI ence to a null hypothesis is common between rerandomisation ~ s~atistics and parametric statis.M l J t h t m U l M . Fisher's. negation. The null model for : shape is the distribution in ~ Kendall's shape space that I arises from landmarks that are < ~ II = = = = = = = = = . the number that a function rule : generates as output for a count~ ing number n. is a propo~ition that is the logical opposite of p.lIO scale data. which have length one. if P is true. NOT p is false. I • nuisance parameters ~ parameters of a model that : must be fit but that are not of ~ interest to the investigator. and if. • not. . . Negation takes precedence over other logical operations. For example. served data. NOT p.nc statlsical mfernece to WhICh rerandomisation statistics are an alternative. NOT p is true. logical negation the negation of a logical proposition p. tiCS. • null hypo eslS .p is false.
.: . .oblique sion.oblique coordinates resents a real number. I 5 • numerical analysis R the study of methods for approximation of solutions of ~ .oblique triangle ing an ellipse through 360 0 . It is exactly analogous to the usual assumption of "independent identically I distributed error terms" in con. a triangle that is not a right triabout its minor axis. angle.oblique prism or cylinder ~ a nonright prism or cylinder MR.number line I a line on which each point rep.thematics==================== II .I ventionallinear models (regres.~ . homoge.p . I a coordinate system in which I the axes are not perpendicular.: . ~ at an angle that is not a mul. x is called the numerator and y is called the denominator.number theory the study of integers.: not equal to zero SIS.I cular normal noise of the same I variance in the original digitising plane or space and I drawn from a single.numerator in the fraction x/y. a line that has a definite slope problems including error analy.oblique line various classes of mathematical . I .oblique angle : an angle that is not 900 . . ANOVA) .null set a set with nothing in it .I neous population. tiple of 90 degrees.II null set I oblique triangle 111 *================ distributed by independent cir. : . f.oblate spheroid an ellipsoid produced by rotat.
If the probability of function if f(x) =f( x) for all x. I all equally likely. pose an experiment can result in any of n possible outcomes. which is the number of favourable outI comes divided by the number I of unfavourable outcomes. and the odds in I favour of winning are (k/n)/ «nk)/n) = k/(nk). an event is p. and that k of the outcomes result in a "win" • octahedron and nk result in a "loss. favour of an event are q. If the odds in dinate planes. ~. then • odd function I the probability of the event is a function f(x) is called an odd I q/(l +q).112 =========* • observational study c. the odds in • odd node I favour of the event are p/(la node with an odd number of ~ p) and the odds against the event are (lp)/p.Note that odds are not syn• octant onymous with probability. arcs . but anyone of the 8 portions of .and forth.f. controlled experiment. the chance of not winning is (nk)/n. suptuse angle. the two can be converted back space dtermined by the 3 coor. I • I odd number an integer that is not diVisible by 2." a polyhedron with 8 faces. • octagon an eightsided polygon.. • obtuse angle an angle whose measure is greater than 90 but less than 180 degrees. For example. II = = = = = = = M l l t h m u J n e s . • odds I the odds in favour of an event I is the ratio of the probability that the event occurs to the • obtuse triangle probability that the event does a triangle that contains an obI not occur. I Then the chance of winning is kin.
~ the rule for using operations on there is some a in A such I numbers. XOR. two rays with a common end: point that form a line I : _ opposite side ~ (of an angle of a triangle) The I side that is not a side of the : angle. .order of operations onto B if for every b in B. the value may range from zero to infinity. smallest to largest or largest to point. first parentheses. : smallest.opposIte rays .OR disjunction. That is.order ing that uses one vanishing ~ to place numbers in order from . this quantity is termed the odds ratio. it is false.tics======= II .f.: .odds ratio an alternative characterisation of the parameter 'p' for a binomial process is the ratio of . C. I : . a segment onto . an interval that does not include : its two endpoints. (p OR) q is a ~ proposition that is true if p is . exclusive disjunc: tion. This relates to a possible view of a binomial process as the com bined activity of two Poisson processes with a limit upon total count for the two processes combined. 113 *================ ~ .°PPOSI e laces I f: th l i ' all I I : aces at e m par e panes I " .onedimensional having length. MIIthem4. ~ "t C . a function f is said to map A : .onepoint perspective I a method of perspective draw. a ray. then multiplif(a) =b. f t h e InCIdences 0 the two alternatives : p/( Ip) . . but no width. examples: a line.II oddsmtio I orderofopemtions . : otherwise.one to one a function f is said to be one to one if f(x) =f(y) implies that x=y. I . If p and q are two propositions.open interval . logical ~ disjunction I an operation on two logical ~ propositions. . are false. (p ~ OR) is true unless both p and q . . : then exponents. true or if q is true (or both).
• ordination a representation of objects with respect to one or more coordinate axes. Biplots combine an ordination of speci I I I I I I II = = = = = = = M s J t b m u J t i c s . a variable whose possible valufs are {straight. warm. then addition and subtraction. These characteristics may arise from categorical rating scales. y. equalto or greater as compared with other data on the ordinal scale. hot}. 3. Arithmetic with the possible values of an ordinal variable does not necessarily make sense. long}. y) • ordered pair rule a rule that uses ordered pairs to describe a transformation.114 orderedpair lordinatiml I • II =========* cation and division. plotting objects according to their scores on the first two principal component axes provides the twodimensional ordination best summarising the total variability of the objects in the original sample space. z) • ordinal scale a measurement type for which the relative values of data are defined solely in terms of being lesser. There are many kinds of ordinations depending upon the goals of the ordination and criteria used.. but it does make sense to say that one possible value is larger than another. {cold. or from converting interval scale data to become ranked data. ordinal variable a variable whose possible values have a natural order. or {O. California. . • ordered triple the three numbers (called coordinates) that are used to identify a point in space. medium. 1. For example.. In contrast. written (x. I I I I I I I I • ordinate the ycoordinate of a point in the plane. Montana. written (x. y) ??(x + h. such as {short. 2. the ordered pair rule (x. }. • ordered pair the two numbers that (called coordinates) are used to identify a point in a plane. curly} or {Arizona. y + k) describes a translation horiwntally by h units and vertically by k units. New York} would not naturally be ordinal. For example.
A I • I I I I orthogonal superimposition a superimposition using only transformations that are all Euclidean similarities. the direcI matrix P. how Ns posi. involve only translation. ===========1~1~5 ~ • orthogonal at right angles. Orthomeans "straight. e. front." MMhemlJtics===================== II .. reflection. I • I I B c orthographic drawing a drawing of the top. i. being "at right angles" is • orientation I defined relative to a symmetric in an image change..0).and yare orthogonal with reI spect to P if xtpy=O.II orientation I orthographic drawing . Principal lates to C's).I tion relates to B's and B's re. two vectors x named go (i. I ance matrix . and. rotation. mens and an ordination ofvariables. and ures ~ relative warps are orthogonal with respect to both bending • origin the point in a coordinate plane energy and the sample covariwith coordinates (0. either clockwise warps are orthogonal with reor counterclockwise for fig~ spect to bending energy. possibly. I • orthic triangle the triangle whose vertices are the feet of the altitudes of a given triangle.e. and right side views of a solid that preserves their sizes and shapes.. In linear algebra. scaling." the views of an orthographic drawing show the faces of a solid as if you were viewing them "headon. such as the bendingtion in which the points energy matrix. • orthocenter the point of concurrency of a triangle's three altitudes (or of the lines containing the altitudes).
before any rerandomisation . a light ray) and the surface it collides with (a cushion. • outlier an outlier is an observation that is many SD's from the mean. • outgoing angle the angle formed between the path of a rebounding object (a billiard ball. • overlapping angles property I the property that. discarding outliers can cause one to underestimate the true variability of the measurement process.orthommnal I operlapping triangles I II • orthonormal a set of vectors is orthonormal if each has length unity and all pairs are orthogonal with respect to some relevant matrix.angles are congruent. Otherwise. overlapping triangles triangles that share a side or angle I • I I I II =======MRrh_tics .. I I I fied. the ~apital letter S. The out. such as the identity matrix. It is sometimes tempting tQ discard outliers. if two angles have the same vertex • outcome space the outcome space is the set of ~ and overlap so that the all possible outcomes of a given I nonoverlapping parts of the random experiment. P. A matrix is orthogonal if its rows (columns) are orthonormal as a set of vectors. if two line segments on the same line overlap so that the nonoverlapping parts are congruent. then the line segments are congruent. • overlapping segments I • outcome value the value of the test statistic for the data as initially observed. a mirror).116 . then the come space is often denoted by angles are congruent. and the outlier is determined to be spurious. but this is imprudent unless the cause of the outlier can be identi I I property the property that..
the Fdistribution ~ used to test decompositions I of variance has two par am MRt'nu======= II . two or more coplanar lines : that have no points in com~ mon or are identical (eg. • pandigital a decimal integer is called I pandigital if it contains each of the digits·from 0 to 9. ~ • parallel lines . : lying in the same plane and not I . rays. I • parameter : in general. a deciI mal) indexing a function. 117 *================ ~ • parallel . ~ • paraboloid a paraboloid of revolution is a I surface of revolution produced by . : mtersectmg. : a quadrilateral in which both ~ pairs of opposite sides are parI aIle!. or line segments). the I same line) I • parallel planes planes that have no points in • pandiagonal magic square a magic square in which all the : common broken diagonals as well as the ~ • parallelepiped main diagonals add up to the I a prism whose bases are magic constant. . • parallelogram rotating a parabola about its axis. (lines. paraIleograms. • palindromic a positive integer is said to be palindromic with respect to a base b if its representation in base b reads the same from left to right as from right to left. a parameter is a ~ number (an integer.II palindrome I parameter • palindrome a positive integer whose digits read the same forward and backwards. For : instance.
I in the space of the partial metric parameters.tations usually do not involve pressed. For the nonunianalysis are expressed.I the inversion of . the coef• parameter I ficients for the rotation are the a numerical property of a popu. the zthe coordinates of x and y as I coordinates. II ====================MlJthmuJrics . In I assessing relationships among morphometrics.I • partial warp scores ences or correlations.118 parameter I partial warp srores !I ========* eters. then to • parametric equations two equations which express I the ycoordinates and. Partial Least Squares which must be estimated to I analyses the covariances beaccount for differences not of ~ tween the sets of variables particular scientific interest. called the pa.principal warps.I quantities that characterise tion is confronted with that I the location of each specimen data.I the uniform partial warps are produced by special. Coefficients for separate functions of a common variable.I partial Least Squares is a muldom for the two variances tivariate statistical method for whose ratio is being tested. there are I two or more sets of variables four main kinds of param. both integers: the I • partial least squares counts of the degrees of free. for threedimensional data. Their compuwhich landmark shape is ex. in the various sets. in which I around the Procrustes mean the findings of the statistical configuration.measured on the same entieters: nuisance parameters. rather than optimizing linear the geometric parameters. such as its mean.warps. ties. and another set of geo. form partial . such as mean differ. by partial warp scores are the which biological interpreta. They are a rotation of partial warp scores or I the Procrustes residuals Procrustes residuals. applied first to the xcoordinates of the lation. eters. I Procrustes residuals. rameter that is usually time. such as warps. statistical param. I combinations of variables in such as shape coordinates.
. ~ • pell number . defmed by the re• partition a partition of an event B is a ~ currence collection of events {AI. Geometrically. I • pedal triangle . • f matrIX .=l. the pedal triangle of a point P with respect to a triangle ABC I is the triangle whose vertices are the feet of the perpendiculars dropped from P to the I sides of triangle ABC. P. . MAthem4rics======= II . AjAk = {} unless j = k. ~ 119 *========= • patterns in fractals recur. A2. 2' in the collection are disjoint. . } such that the events . S. Ipayo f ' . 1 +Pn . . a way of representing what each player in a game wins or loses. sive rule : a rule used to fmd terms in num~ ber sequences using recursion. 2. Algebraically.~ a ISsided polygon haust B). • pentagonal number be the entire outcome space S. the one for uniform shape change. Except for the very largestscale partial warp. it is assumed to . That is.. : a number of the form n(3nl)j ~ 2. the nth term in the sequence 0. A3. . . Po=O. not specified. they have an approximate location and an approximate scale. and B = Al U A2 ~ • pentagon U A3 U . • Pascal's triangle a triangular array of binomial ~ • pentomino coefficients. . a fivesided polygon.. 12. and Pn =2Pn . .II partial warps I pentomino • partial warps partial warps are an auxiliary structure for the interpretation of shape changes and shape variation in sets of landmarks. a fivesquare polyomino. I as a function of his and his ~ opponent's strategies. : 1. : • pentadecagon and their union is B (they ex. If the event B is . the partial warps are eigenvectors of the bending energy matrix that describes the net local information in a deformation along each coordinate axis.. partial warps are an orthonormal basis for a space tangent to Kendall's shape space.
For example.. an integer is a perfect power there are n2.120 percentile Ipennutation =================* II ~ . .. m and n are integers and n> 1. 28 is perment in an ordering... • the set in some order. If the • perfect number set has n things in 'it... • ond. The pth percentile of a random vari able is the smallest number such that the chance that the random variable is no larger than it is at least p%. etc.f.. • • perfect square I an integer is a perfect square if it is of the form m 2 where m is an integer. quantile.. • percentile the pth percentile of a list is the smallest number such that at least p% of the numbers in the list are no larger than it. there are a positive integer that is equal I nl different orderings of its to the sum of its proper divi• elements. C.. for the sec4+7+14. • • • • B Closed (Periodic) Cubic S5pline A c: • perfect cube an integer is a perfect cube if it • • permutation is of the form m 3 where m is an a permutation of a set is an arrangement of the elements of integer. for the third.. By the fundamental II = = = = = = = = = = M l l t b m u J t i & s . For the first elesors. there remain n1 pos• perfect power • sible choices. there are fect because 28 = 1 + 2 + • n possible choices.. • • perimeter of a polygon • the sum of the lengths of the sides of the polygon • periodic curve • a curve that repeats in a regu• lar pattern. and for the if it is of the form mn where • nth element of the ordering. • there is a single choice remammg.
ships realistically in a draw: ing. : pendent variable.f. Pitman who described : this test. and the PITMAN • perpendicular planes planes in which any two inter. .. • perpendicular bisector (of a line segment) A line that ~ • piecewise function divides the line segment into : a function consisting of two or two congruent parts (bisects ~ more equations. ~ exact rerandomisation test in I which the test statistic is the Q : difference of means of two OR ~ samples of univariate interP B I valscale data. this is one of the earliest in: stances of an exact test.~ permutation test. feeling of depth . combinations.==========~1~2~1 rule of counting. N • C pi : written 7t the ratio C[D where ~ C is the circumference and D is I the diameter of a circle. specified intervals of the indeit. defined for it) and is also perpendicular to . This is denoted np k' the number of permutations of n things taken k at a time. or lines that ~ named after the statistician form a 90 degree angle . I • perpendicular lines : • Pitman permutation test 2 segments.. E. ~ • perspective . I. . x(nk+ 1) = n!j(nk)1. • perpendicular intersecting at right angles. by drawing objects ~ smaller as they recede into I the distance. . one in each plane. the number of orderings of length k one can form from n> = k things is nx(n1)x(n2)x . MIIthematics================== II . Similarly.14159265359 . secting lines. rays. the total number of sequences is thus nx(n1)x(n2)x . .J . • perspective drawing : a technique of representing ~ threedimensional relation. xl. in 1937..II perpendicular I Pitman permutation. C. ~ 3. An form a right angle.
e. this is the essence of a blind experiment. if the variable X is plotted on the horizontal axis and the variable Y is plotted on the vertical axis. • placebo a "dummy" treatment that has no pharmacological effect. It has no size. Subjects given a placebo as a painkiller report statistically significant reductions in pain in randomised experiments that compare them with subjects who receive no treatment at all. • plane an undefined term in most deductive systems. • plane figure a set bf pO'ints that are on a plane plane geometry the study of twodimensional figures in a plane I I • I I I I I I I I I I • plane section the intersection of a figure with a plane • point an undefined term in most deductive systems. is called the placebo effect. I • I I I I I II =======Mnthmultics . A plane has length and width but no thickness and is therefore considered twodimensional. mean of Y). Administering a placebo to the control group is thus important in experiments with human subjects. the point of averages has coordinates (mean of X.g. which has no direct biochemical effect. You can think of geometric figures as sets of points. A flat surfa~e that extends infinitely. only location and is therefore considered zerodimensional. In a Cartesian coordinate system.. For example. This very real psychological effect of a placebo. point of averages in a scatterplot. the point whose coordinates are the arithmetic means of the corresponding variables. a sugar pill. a point's location is represented by a pair of numbers (xJ y).122 pixel I point ofaverages I I !I =================* • pixel small dot of color that makes up computer'and TV screens • placebo effect the belief or knowledge that one is being treated can itself have an effect that confounds with the real effect of the treatment.
the probability is zero. ... . • Poisson distribution the Poisson distribution is a discrete probability distribution that depends on one paranleter.vents occur independently in some COI1tinuwn (in many applications. k = 0.II pointofconcurrency I polyhedron • point of concurrency the point at which more than two concurrent lines. A line Mathematics===================== I . The corresponding distribution of intervals between events is an exponential distribution. • Poisson process a process whereby t. If X is a random variable with the Poisson distribution with parameter m. each of which intersects exactly two others at endpoints forming the polygon'S angles. or rays intersect. 1. The expected value the Poisson distribution with parameter m is 111. where E is the base of the narurallogaritlun and ! is the factorial function. m. such that the overall density (rate) is statistically constant but that it is impossible to I 123 *========= I I I I I I I improve any prediction of the position (time) of the next event by reference to the detail of any number of preceding observations. its direction polygon a closed planar geometric figure consisting of line segments (tlle sides). • polarity of a variable the positivity or negativity of a variable. • point of tangency (of a circle) The single point where a tangent line touches a circle. each of which is a face. line segments. Each point of intersection is a vertex of the polygon. and the standard error of the Poisson distribution with parameter m is m 1/2 . then the probability that X = k is Em x mkjk!.2. time). The conventional example of a Poisson processes is concerned with occurence of radioactive emissions in a substantial sample of radioactive with a halflife very much longer than the total observation period. For all other values of k. I I I • I I I I I • I I I polygonal region the union of a polygon and its lntenor • polyhedron a solid whose surface consists of polygons and tlleir interiors.
values of a variable for a popu.. the percentage of units in a I population that possess a specified property. C. the population percenta sample subset are intended I age is the population mean for to be applied. I II = = = = = = = M J J t__#U .polyomino a planar figure consisting 0 f I congruent squares joined edge.i1W I population stand4rd deviation I II segment where two faces intersect is an edge. sample standard numerical' population.population standard deviation population is essentially inforI the standard deviation of the mal.. For ex.I parameter. the percentage of a given collection I of registered voters who are registered as Republicans. The population I mean is a parameter." and each lmit that does not possess the property is labelled . The population a population of zeros and ones. the population mean of a box of numb.f. process of applying inferences based upon the sample to the .population I with "0. C. If each unit that possesses the I property is labelled with "1.deviation. c. ample.ered tickets is the mean of the list comprised of all the numbers on I all the tickets.: . : sample mean. not the mean of the numbers in a I a statistic.population mean lation.The population percentage is a ber the sample. This is a parameter. . For example. sample percentrandomisation statistics the age.f.124 P01. In re.f. A point of intersection of three or more edges is a vertex. from statistical examination of I that is." the population pera definable set of individual I centage is the same as the mean units to which the findings of that list of zeros and ones.population percentage toedge. will generally much outnum.
II populatum I preshape space
• population a collection of units being studied. Units can be people, places, objects, epochs, drugs, procedures, or many other things. Much of statistics is concerned with estimating numerical properties (parameters) of an entire population from a random sample of units from the population. • post hoc ergo propter hoc after this, therefore because of this. A fallacy of logic known since classical times: inferring a causal relation from correlation. Don't do this at home! • postulates premises in a deductive system accepted without proof. • power this is the probability that a statistical test will detect a defined pattern in data and declare the extent of the pattern as showing statistical significance. power is related to type2 error by the simple formula: power = (Ibeta) ; the motive for this redefinition is so that an increase in value for power shall represent imI
125
*=================
I
provement of performance of a statistical test. practical number a practical number is a positive integer m such that every natural munber n not exceeding m is a sum of distinct divisors of
I •
I I I
I
I
m. • precision the closeness of repeated measurements to the same value. • preform space the space corresponding to centered objects, i. e., differences in location have been removed. It is of k(PI) dimensions. • preimage the original object that is reflected premises (in a deductive sy~tem) Statements (including undefined terms, definitions, properties of algebra and equality, postulates, and theorems) used to prove further conclusions.
I
I
I •
I I
I
I
I •
I I
preshape space the space corresponding to figures that have been centered and scaled but not rotated to alignment. It is of k(Pl)1 dimensions.
Mathematics===================== II
""12""6==========,,,,p"=·nu;aCie I p1incipal components atlalysis
I
• prima facie latin for "at tirst glance." "On the face of it." Prima facie evidence for something is information that at first glance supports the conclusion. On closer examination, that might not be true; there could be another explanation for the evidence. • prime a prime number is an integer larger than 1 whose only positive divisors are 1 and itself. • prime factorisation the unique set of factors of a number, all of which are prime numbers. • primitive pythagorean triangle a right triangle whose sides are relatively prime integers.
I I
I
I
I
I I
I
I I
I
its effect on a circle or sphere. An affine transformation takes circles into ellipses. The principal axes of the shape change are the directions of the diameters of the circle that are mapped into the major and minor axes of the ellipse. The principal strains of the change are the ratios of the lengths of the axes of the ellipse to the diameter of the circle. In the case of the tetrahedron, there are three principal axes, the axes of the ellipsoid into which a sphen: is deformed. One has the greatest principal strain (ratio of axis length to diameter of sphere), one the least, and there is a third perpendicular to both, having an intermediate principal strain.
principal components analysis • primitive root of unity the complex number z is a the eigenanalysis of the primitive nth root of unity if ~ sample covariance matrix. Zll = 1 but Zk is not equal to 1 for ; Principal components (PC's) any positive integer k less than can be defined as the set of I vectors that are orthogonal n. both with respect to the iden• principal axes and strains tity matrix and the sample a change of one triangle into I covariance matrix. They can another, or of one tetrahedron also be defined sequentially: into another, can be modelled the first is the linear combias an affine transformation I nation with the largest variwhich can be parameterised by
I •
II ====================MiJthemllhcs
II principal wa1'"Ps I prism..
12 7
ance of all those with coeffi I set of p 2D landm~irks (p4 for cients summing in square to I 3D data) they form a finite seI; the second has the largest ries . Together with the uniform variance (when normalised I terms, the partial warps, which that way) of all that are , are projections (shadows) of the uncorrelated with the first principal warps, supply an orone; etc. One way to compute thonormal basis for a space that principal components is to use I is tangent to Kendall's shape a singular value decomposi space in the vicinity of a mean tion. Relative warps are prin form. cipal components of partial I • prism warp scores. There is a lot to I a polyhedron with two congrube said about PC's; see any of ent polygons in parallel planes the colored books. as bases. Line segments (lateral • principal warps I edges) connect the correspondprincipal warps are ing bases to form lateral faces, eigenfunctions of the bending which are parallelograms. An energy matrix interpreted as I altitude is a line segment beactual warped surfaces (thin I tween, and perpendicular to, the plate splines) over the picture planes of the bases. The height of the original landmark con I is the length of an altitude. If figuration. Principal warps are I the lateral edges are perpenlike the harmonics in a Fourier dicular to the bases, the prism analysis (for circular shape) or is a right prism; otherwise it is Legendre polynomials (for lin I oblique. ear shape) in that together they decompose the relation of any sample shape to the sample av , erage shape as a unique swumation of multiples of eigenfunctions of bending en , ergy. They differ from these more familiar analogues in that there are only p3 of them for a I
MRthemilties=================== II
Only 5 4/36 continuous random variables 6 5/36 I 7 6/36 II = = . = = = = = = = M a t " . The I mously. That is. For example. However.. any rule for I the real numbers . the terms lated as the area under a curve sometimes are used synonyover that range of values. number between zero and I • probability distribution 100%.2 1/36 tion of a random variable with 3 2/36 3/36 a standard normal distribu. . The meaning (interpre. I variable takes each of its posthere is a function f(x) such sible values. I s the probability density func. The probability districurve is the probability den. .t i c s .bution function.probability I probabilitydistrilmtilm " • probability I have probability density functhe probability of an event is a I nons. . .I by the chance that the random tinuous random variable.I bution of a discrete random sity function of the random variable can be characterised variable. f is the probability density can be written as a table: P(S=s) function of X.the probability distribution of tation) of probability is the sub.. (The subassigning probabilities to events sets have to satisfy some techhas to satisfy the axioms of ~ nical conditions that are not probability. P(a<= X <=b) = I total number of spots S show(area under f between a and ing on the roll of two fair dice b). For example. the that for every pair of numbers probability distribution of the a<=b.) I The probability distribution of • probability density function a random variable is comthe chance that a continuous pletely characterised by the random variable is in any I cumulative probability distrirange of values can be calcu.I a random variable specifies ject of theories of probability..""12.I 4 tion is the normal curve. if X is a con. I the chance that the variable which differ in their interpretatakes a value in any subset of tions.8========== . . important for this course.
~ I : ~ . a probability is a number .go.45 0 . any subset of the sample space IS an event.9 8 5/36 9 4/36 10 3/36 11 2/36 12 1/36 The probability distribution of a continuous random variable can be characterised by its probability density ftillction. 0 . that measures how strongly : we believe an event will occur. the probability of an event is I the limit of the percentage of times that the event occurs in repeated.. • probability histogram a probability histogram for a ran d om varia bl e is ana Iogous to a histogram of data. theories of I a theory of probability is a way : of assigning meaning to prob~ ability statements such as "the I chance that a thumbtack lands ~ pointup is 2/3. There are several . According to the : subjective theory of probabil~ ~ ity.. which is the set of conseq uences of the axioms of ~ probability. common theories of probabilI ity. • probability events the set of all possible outcomes of an experiment is the sample space.1 0 . world of observation and ex~ periment.ple a sample drawn from a population using a random mechanism so that every element of the population has a known chance of ending up in the sample.4 f o 1 _oI.. but instead of plotting the area of the bins proportional to the relative frequency of observations in the class interval. cumstances. quency theory of probability.2 0. 1ro~ L. a theory of probability connects I the mathematics of probabilI ity.25 0. independent trials ~ under essentially the same cir. with the real . According to the fre.05 .15 0. one plots the area of the bins proportional to the probability that the random variable is in the class interval. !1 0.II probability wmts I probability} th~0""if==========1""2." That is. : ~ • probability sarr.x ~ • probability. ~ The number is on a scale of Mslt'tics======= " .3 0.
1905). and its detractors.=13=0========'1'== .: Mosier (1939). ana (for example. (by leastsquares) superim. the best choice of ~ (Boas. then imposition methods by Hudey the chance of each outcome is I and Cattell. 7:258262.. Cole (1996) reports posed configurations at centroid that Franz Boas in 1905 sugsize. The 100%jn.. experiment has n possible out.I most subsequent morphometing that we are completely I ric analyses . Sci. its proing a direct rotation to test an ponents. developer of these methods.However.= = = = = = M i J t h m u l r i e s . I ods for estimating nuisance According to the theory of : parameters of the Euclidean equally likely outcomes.the square root I ern workers have often cited · of the sum of squared differ.I The adjective "Procrustes" comes. Cole further points out that one of Boas' consensus configuration for .I tims to fit a bed and was first comes should occur preferenI used in the context of supertially to any of the others. This is the distance that I gested the "method of least (ordinary defines the metric for Kendall's differences" shape space. Each of these theoI Procrustes program: producries has its limitations. as the earliest known the landmarks in two optimally. a psychomeences benveen the positions of ~ trician. if an ~ similarity transformations. students extended the method 11 = . by I refers to the Greek giant who symmetry) there is no reason would stretch or shorten victhat any of the n possible out.ocrusteidistance Iprocrustes methods 11 0% to 100%. with 0% indicat. Modapproximately . and 100% I • procrustes methods indicating that we are coma term for leastsquares methpletely sure that it will occur. 1962. Procrustes analysis) as a I means of comparing homolo• procrustes mean the shape that has the least I gous points to address obvious problems with the stansummed squared Procrustes distance to all the configurations I dard pointline registrations of a sample. sure it won't occur. hypothesised factor structure. • procrustes distance I Behav.
• procrustes residuals the set of vectors connecting the landmarks of a specimen to corresponding landmarks in the consensus configuration after a Procrustes fit. then b = a). The latter being essentially a Generalized Procrustes Analysis. The sum of squared lengths of these vectors is approximately the squared Procrustes distance between the specimen and the consensus in Kendall's shape space. to I • 131 *================= I I procrustes superimposition the construction of a twoform superimpOSitIOn by least squares using orthogonal or affine transformations. and transitive property (if a = band b = c.two ratios.II procrustes residttals I proportional the construction of mean configurations from the superimposition of multiple specimens using either the standard registrations of Boas' method (Phelps. The partial warp scores are an orthogonal rotation of the full set of these residuals. symmetric property (if a = b. • proof a sequence of justified conclusions used to prove the validity of an ifthen statement • proper divisor the integer d is a proper divisor of the integer n if O<d<n and d is a divisor of n. 1932). • properties of equality reflexive property (a = a). then a = c). tered at the corresponding • proportional landmark of the Procrustes lone of four numbers that form mean shape. • proper subset a subset that doesn't include everything in its parent set. a scatter of all the I a statement of equality between Procrustes residuals each cen. I • pronic number a number of the form n(n+ 1) . I I I I I I • procrustes scatter I a collection of forms all superimposed by ordinary orthogonal Procrustes fit over one I single consensus configuration that is their Procrustes • proportion mean. I a true proportion MRthem4tics=================== II .
under the null hypothesis. proportWnality Ippalue I I II • proportionality a relationship described by a constant ratio. • protractor a tool used to measure the size of an angle if'. • pvalue suppose we have a family of hypothesis tests of a null hypothesis that let us test the hypothesis at any significance level p between 0 and 100% we choose. theorems. degrees. let xp be the smallest number such that. I I I I process. Successive pseudorandom data are produced by a fLxed calculation process acting upon preceding data from the pseudorandom sequence. P( X < = x ) > = p.132 =======~. • pseudorandom a source of data which is effectively unpredictable although generated by a determinate I I I I I II ==========MRth_ties . let X be a test statistic. logical proposition a logical proposition is a statement that can be either true or false. The P value of the null hypothesis given the data is the smallest significance level p for which any of the tests would have rejected the null hypothesis.. and properties to show that a relationship is true for all numbers or specific set of figures. and for p between 0 and 100%. For example. • proposition. I I I I I • prove use logical arguments. Then for any p between o and 100%. If we observed X = x. "the sun is shining in Berkeley right now" is a proposition. To start the sequence it is necessary to decide arbitrarily upon a first datum. definitions. the Pvalue of the null hypothesis given the data would be the smallest p such that x < xp. For example. which is termed the seed value . the rule reject the null hypothesis if X < xp tests the null hypothesis at significance level p.
The altitude is the line segment from the vertex ending at and perpendicular to the plane of the base. If the line segment connecting the vertex to the center of the base is perpendicular to the base. This appeared originally ~ at the end of many of Euclid's I propositions. triple is a multiple. then the . Their corresponding sides are proportional. • pythagorean fractal similar figures figures that have the same shape but not necessarily the same size. 133 . • qed : ((quod erat demonstrandum)) ~ (Latin) This stems from medi. . If the three integers I have no common integer facI tors. otherwise it is oblique. the sum of the : done" It appears in Latin MAthmuJti&s======= : . strated" to the Latin phrase : above. signifying that he ~ had proved what he set out to : prove.==============~ I squares of the lengths of the .qef I II . • pythagorean theorem I ((quod erat faciendum)) is the the theorem that says that. I : • pythagorean triangle ~ a right triangle whose sides are I integers. . • pythagorean triple : three positive integers with I : the property that the sum of I the squares of two of the in: tegers equals the square of the third. The height is the length of the altitude. legs equals the square of the : length of the hypotenuse. If the three integers have I a common factor. eval translators' habitual ten: dency of translating the Greek ~ for "this was to be demon. then the pyramid is right. in a ~ latin for "which was to be right triangle.II pyramid I qef • pyramid a polyhedron with a polygon base and line segments connecting the vertices of the base with a single point (the vertex of the pyramid) that is not coplanar with the base. then the triple is primi: tive.
quadrilateral a foursided polygon . equation amounts of money. genders. temperatures. • quadrangular pyramid a pyranlid whose base is a quadrilateral. weights. C.. I a variable that takes numerical ax2 +bx+c=0.and categorical vari• quadratic equation I able. etc. values for which arithmetic I makes sense. etc. nasquare free tionalities.. I • • quadrinomial • quadrant : an algebraic expression consistanyone of the four portions of I ing of 4 terms. for example. • quadratfrie I such as colours. whose values are adjectives. • quadratic term of an I counts. I • quadric surface the graph of a second degree equation in three variables. I I quadrature the quadrature of a geometric figure is the determination of its area. • quadrangular prism a prism whose base is a quadrilateral.f. That IS. • quadrangle a closed broken line in the plane consisting of 4 line segments. For the term AX2 in a quadratic I some variables that take nuequation I merical values. I • I quadric curve the graph of a second degree equation in two variables. an equation of the form f(x) =0 where f(x) is a second degree I • quantitative variable polynomial. qw:ulrangle I quantitative l1ariable 1/ I • translations of Euclid's works signifying that he had demonstrated what he had set out to demonstrated. such variables are not II = = = = = = = M 1 I t__ries .134 . quantitative variable . arithmetic with those values does not make sense. the plane into wruch the plane : • qualitative variable I is divided by the coordinate a qualitative variable is one axes.
people. ..MR. and at least 1/4 of the numbers in the list are no smaller than it. or pounds. Call that integer m. . : The mth element of the sorted 1 list is the upper quartile. .. first sort the list into increasing order.radical axis : the locus of points of equal 1 • : power Wit h respect to two 1 circle. Call that integer k. The kth element of the sorted list is the lower quartile. Quantitative variables typically have units of measurement. ~ Quota Sampling. The first or lower quartile (LQ) of . Find the smallest integer ~ that is at least as large as the 1 number of entries in the list ~ times 3/4.quartiles there are three quartiles.thmulties=================== II . Find the smallest integer that is at least as big as the number of entries in the list divided by four. One radian .qUlnnc po ynoml 1 . . ~ a polynoffilal of degree 5.r : radius ~ _ radian 1 a unit of angular measurement such that there are 2 pi radians in a complete circle. The lth element . .quotient 1 the result of a division. a hst IS a number (not necessarily a number in the list) such that at least 1/4 of the numbers in the list are no larger than it. . The second quartile is the median. To find the quartiles. For example. : . I 'al .IllJUllrtic polynomial I radical axis quantitative. The third or upper quartile (UQ) is a number such that at least 3/4 of the entries in the list are no larger than it.quartic polynomial a polynomial of degree 4. : . big as the number of entries : in the list divided by two. Find the small~ 135 *================ est integer that is at least as . is approximately 57. Call ~ that integer 1._R · I rotation . One radian ~ = ISOjpi degrees. and at least 3/4 of the numbers in the list are no smaller than it. adding and subtracting social security numbers does not make sense.30. of the sorted list is the me: dian. such as inches. .
\0'" \ I I each time a measurement is made. J J '\. that is. which is in general different I I I I I I I I • random sample a random sample is a sample whose members are chosen at random from a given population in such a way that the chance of obtaining any particular sample can be computed. and a random error. I dictable. "haphazard. the length of that segment • radix point the generalisation of decimal point to bases of numeration other than base 10. plural form of radius \ 0 A '"'." which does not I necessarily imply longterm • radii regularity. peated trials is predictable. which can often be broken down into two components: a bias or systematic error. which affects all measurements the same way. but for which the long• radical center run relative frequency of outthe radical center of three circles I comes of different types in reis the common point of. • random error all measurements are subject to error. outcome is not perfectly pre\. often denoted n. Note interesection of the radical axes that "random" is different from of each pair of circles.\. . The number of units in the sample is called the sample size. drawing all n objects in the sample at once (a random sample without replace II ===================MJJt__ncs . Random samples can be drawn with or without replacing objects between draws. and behaves like a number drawn with replacement from a box of numbered tickets whose average is zero.136 radical center I random sample II ========*  random experiment an experiment or trial whose . f . The number of units in the population often is denoted N. B '\ \ I I . I • • radius the segment whose endpoints are any point on a circle or sphere and its center.
. or drawing the objects ~ 1 to the outcome {T. Th e term ran om· l' I 1 h i d I pen menta umts. more than once.randomisation distributuple of units in the popula. t h"IS IS usuata was samp e f not h e N umts m t e popu.: .~ tion tion is equally likely.111 llndomVilriable I mndomislltiunset . measureable characteristic of dom sample without replace. T. T}. o 137 ment). replacement). The num. H}. ca e 1 a slm p e random ~ and repeated measures to exsamp e. any .: the procedure by which the . : foreseeable possibilty of any any given member of the ~ systematic relationship bepopulation can be in the ~ tween the data and any sample at most once. subset of I d i d . A ran. H}. a collection of values of the test : statistic obtained by undertak. in accordance tion can occur in the sample ~ with the experimental design. randomisations of data within it assigns the number 0 to the : the constraints of the experioutcome {T.rand " ber of heads showing when the ~ the collection of possible recoins land is a random variable: . . ment in which.: . ment. For example. I 11 d O l d : penmenta umts to groups. H. on a stnct y samp e WIt rep acement e: random basis. T. In a random I such that there should be no sample without replacement. In a random ~ _ randomisation sample with replacement. the process of arranging for given member of the popula. and the number 3 to the draws (a random sample with ~ outcome {H.random variable ~ ing a number of rea random variable is an assignI randomisations of the actual ment of numbers to possible ~ data within the randomisation outcomes of a random experi: set. the number ~ mental design 0 0 0 0 0 0 0 MIIth_tics================== /I . the one at a time. notes a random sample drawn I in such a way that every n.: datacollection. every h .: 11 d bOO 11 1 k 1 1 I a Y arrange ' asslgnmg exY 1 anon IS equa y 1 e y IS a so . number to to the outcome {T: in the population between : H. consider I olDlsabon set tossing three coins. replacing them . H}.
N) within the set. ratio of similitude the simplest form ratio of the measures of corresponding parts of similar figures. we could write an identifying number for each subject on a slip of paper. and draw slips without replacement until we have drawn half of them. the range is a single number.8===========. if the outcome value is near to one extreme of the rerandomisation distribution then it may be judged that it is in the extreme tail of the distribution. • rate a ratio where the quantities are . the outcome value value of the test statistic is judged in terms of its relative position within the rerandomisation distribution.1""3. example: 60 miles per hour • ratio the quotient of two numbers... Note that as a statistical term. not a range of numbers. The subjects identified on the slips drawn could then be as I I I I I signed to treatment. • randomised controlled experiment an experiment in which chance is deliberately introduced in assigning subjects to the treatment and control groups. and thus judged to show statistical significance. to be regarded as ordinalscale.. by making the treatment and control groups roughly comparable in all respects but the treatment. and replacing each datum by its rank (1 .ndmnisation test I ratio ofsimilitude II • randomisation test the rationale of a randomisation test involves exploring rerandomisations of the actual data to form the randomisation distribution of values of the test statistic. I I I I I I I I I I I I I • I II = = = = = = = M R .". with reference to a nominal alpha criterion value. For example. and the rest to control. of different kinds. stir up the slips of paper. • ranked data this refers to the practice of taking a set of N data. Randomising the assignment tends to decrease confoimding of the treatment effect with other factors.. • range the range of a set of num bers is the largest value in the set minus the smallest value in the set. t h e m 4 t i c s .
. .recursion to by giving the names of two : the process of generating a sepoints.real numbers . I ous term. the real number x is called ther : eal part of the complex number ~ x+iy where x and y are real and . : plying a rule to obtain any suc.real axis ~ ceeding term from the previthe xaxis of an Argand diagram.rectangle that is the ratio of two integers.rectangular solid . fying a first term and then apthen any point on the ray. to be irrational. A ray is referred . All other real numbers are said : angles. and either continuous or discrete.II ratio scale I recursiveform 139 *================ . a quadrilateral with 4 right. . Such a scale will have a zero point which is meaningful in the sense that it indicates complete absence of the property which the scale measures. The ratio scale may be either unipolar (negative values not meaningful) or bipolar (both positive and negative values meaningful). .real part . i=sqrt( 1).ray I the union of a box and its inteall points on a line that lie on I nor one side of a specified point. I : . . ~ . a formula for the next term of . .rational number a rational number is a number ~ .recompose : put addends or factors of a I given number back together in a way different from the original arrangement or decompoI sition. I . .ratio scale this is a type of measurement scale for which it is meaningful to reason in terms of differences in scores and also in terms of ratios of scores.recursive form rational and irrational numbers.real variable : a variable whose value ranges ~ over the real numbers. . I • I : .reclproca I the reciprocal of the number x : is the number l/x. the ray's endpoint. first the endpoint and ~ quence (or pattern) by speci.
you must subtract 360 degrees from the given angle until there is less than 360 degrees left. Also II ===================MRther_#&s .I gent choice of baseline for the I construction usually does. or mirror line) and are the same distance from the line. fi images of pomts in the Igure reflection notation rm(ABC). In order to get the reference angle. though the intelli.""l40==========~eforencellngle I rejlectionalsymmetry II a sequence given the term before it. There has been some controversery regarding the choice of reference . • reference angle the angle of less than 360 degrees that corresponds to an angle of over 360 degrees. The I reference configuration corresponds to the point of tan. and the orthogonality that characterises the partial warps is with respect to this particular formula for bending energy. The I mean configuration is usually I used as the reference in order to mInImise distortions I caused by this approximation. • • reference configuration I in the context of superimposition methods. this is the con. • reflection image of a figure the set of all of the reflection . the bending energy that goes with them is computed using the geometry of the grand mean shape. The construction of ~ twopoint shape coordinates. I When splines and warps are refine to change a conjecture slightly so that it is true • reflection an isometry under which every point and its image are on opposite sides of a fL\:ed line (the line of reflection. It may be another specimen in the sample but usually I it will be the average (consensus) configuration for a sample.I gency of the linear tangent space used to approximate Kendall's shape space.I figuration to which data are fit. does not involve a reference specimen. I I I I I I I I part of the analysis. \vhich stands for the reflection over line m of figure ABC • • reflectional symmetry the property of a figure that it coincides with its image under at least one reflection.
the two variables measured are the height of the father and the height of the son.I ric object that it is congruent to I itself. figure a figure that shows reflection symmetry I 141 • regression fallacy the regression fallacy is to attribute the regression effect to an external cause. son) pair chosen at random from this I • reflex angle : an angle between 1800 and 3600. I I I ~ I : ~ . If the value of the first variable for that individual is above average. son) pair. regression effect suppose one measures two variabIes for each member of a group of individuals. whether peas or people. the value of the second variable for that individual is likely to be above (below) average. symmetry. : ~ MRth_tics======= II . suppose one measures the heights of fathers and sons. These two variables will tend to have a positive correlation coefficient: fathers who are taller than average tend to have sons who are taller than average. the word comes from the faG: that when measurements of offspring. • regression toward the mean. Consider a (father. and that the correlation coefficient of the variables is positive (negative). That is. but by fewer standard deviations than the first variable is. regression effict called line symmetry or mirror ~ . I • region the tullOn of a figure and its interior • regression a model for predicting one variable from another. : • reflectionsymmetric I . the second observation is likely to be closer to the mean in standard units.II rejlectWnsymmetricM'tre I regres~oward the mmn. the offspring measurements "went back" or regressed towards the mean. For example. were plotted against the same measurements of their parents. Due to Francis Galton. I ~ • reflex polygon : a polygon for which 2 or more I of its sides intersect each other • reflexive property of congruence I the property of every geomet. Each individual is a (father.
I the correlation coefficient times tion of the fathers' heights. Suppose the father's I zontal axis.) I the SD ofY divided by the SD Then the son's height is also of X. together with the standard deviations of the two variables or I their correlation coefficient. a polyhedron whose faces are • regression. but by I a sixsided figure whose sides fewer than 3SD (here the SD is I are of equal length and whose the standard deviation of the angles are of equal measure. regular sion I polygons that meet at all vertilinear regression fits a line to a I ces in exactly the same way.I passes through the point of avage of all the fathers' heights. In other cases..=========MR.• regular pyramid mary.142 regres.. and has slope equal to (The SD is the standard devia. the I • regular polygon rejection region is the set of : a convex polygon whose angles values of the test statistic for I and sides are all congruent which we reject the null hypoth·· I • regular polyhedron eSlS. the regression line height is 3SD above the aver.I II =========i.. If we are regressing the I a pyramid whose base is a regu" variable Y on the variable X. erages.. can be a reasonable summary of a scatterplot if the scatterplot is I roughly footballshaped....thcmancs . it is a poor sum. scatterplot in such a way as to minimise the sum of the I squares of the residuals. sons' heights). The I resulting regression line. ~kely to be above the average • regular hexagon of the sons' heights. linearngnssWn I regularpyramid II group .... and lar polygon and whose vertex if Y is plotted on the vertical forms a segment with the cenaxis and X is plotted on the hori .. In an hypothesis I test using a test statistic. linear regresenclosed by congruent..
fit rotation estimate is the methe axes of the "ellipsoid" oc. Theyare .: are the same.: ments defined using that land. be computed from Procrustes lar to its plane : residuals or from partial warps. which is. _ repeatedmeasures I used to weight shape variation ~ this is a feature of an experiBy the geometric scale of shape : mental design whereby sevI MJJthematU:s======= II . : that do not share a vertex with .: preted as specifying multiples of ~ mark and the other nl landone single transformation. the parameter can be . a comparison of two or more statistical tests. for the same ~ .: ferences between the reference rance. the median of angular difare defined by Procrustes dis.~ some superimposition paramponents of a distribution of : eters in the resistantfit methshapes in a space tangent to ~ ods. the resistantKendall's shape space. For example.~ (of the exterior angle of a tri. in in a geometry in which spheres . sample I an integer all of whose digits size. _ regular tessellation I an edgetoedge tessellation in : . Repeated medians are transformation that can often : insensitive to larger subsets of be usefully drawn out as a thin. analysis.: _ repeated median spective values of power. in terms of the re.~ extremely deviant values than plate spline. as a ~ configuration and the configudirection of shape change about .remote interior angles which tiles are congruent regu. and nominal alpha crite. Each relative warp.: dian of the estimates obtained cupied by the sample of shapes ~ for each landmark. I rion value.relative power ~ the e:A'terior angle. ration being fit of the line segthe mean form. a I marks. ~ a median of medians. can be inter.rep digit experimental design. In a relative warps I simple medians. angle). tmn.II reguJartessellatWn I repeatedmeaJ1t~ ==========14~3 ter of the polygon perpendicu. Relative warps can . Repeated _ relative warps I medians are used to estimate relative warps are principal com.~ differences. the two interior angles lar polygons.
the same sampling unit. II =======Msthem4ries . the difthe sample is representative.g. ternative arrangements of expenmental design is crucial I given data which would be to this definition. an integer consisting onlyofl's. I • resampling stats • replications this is a feature of an experi.tion of the randomisation disI tribution of a defmed test starelevant.144 ~~~~~~~==* replications I resitlfuU II eral observations measured ~ • repunit on a common scale refer to . in the form of an intersame conditions. ~ ferences between a shape and in practical terms. • representative patterns in a sample of units I • residual may reasonably be attributed the deviations of an observed to the population from which : value or vector of values from the sample is drawn. Identification of the relation of the I • rerandomisation individual observations to the I the process of generating al. Examples : the measurement of water I consistent "\vith the experimental design.preter. These using right hand and left hand are the specific area of concern I of this present glossary.par. separatel.this is the name of an educamental design whereby obser. e.I specify montecarloresampling ticular observation within the I of a set of data and accumulasequence of replications is ir. to ensure its prediction by an allometric that a sample is representative I regression expressed in any set almost always means ensuring of shape coordinates. that it is a random sample. measurement of I also known as permutation or reactiontime of an individual : randomisation statistics. Identifica. Ievel at a particular site on several systematicallydefined : • rerandomisation statistics occasions.tional initiative involving the vations on an experimental I use of a programming lanunit are repeated under the I guage. only if I some expectation. allowing the user to tion of the position of a .. I tistic .
imaging. ~ superimposition methods that • residual . or the sum of the residuals from ~ display device. the mea residual plot for a regression . : tors. use medianand repeatedmethe difference between a datum : dianbased estimates of fitting and the value predicted for it by ~ parameters rather than leasta model. is a plot of the residuals from i • resistantfit superimposithe regression against the extion planatory variable. there is a com. such I ting methods. dian is.II residualplot I resultant'flC&tor' 145 • residual plot ~ mean is not resistant. • resolution heteroscedasticity. In linear regression of I squares estimates.: results may provide a simple siduals can be positive (if the ~ description of differences in datum is above the line) or . as nonlinearity and. slide the original vectors so a statistic is said to be resistant ~ that their tails intersect. the regression line must be zero. otherwise. The ~ mon tail. resistantreveal computational errors in i fit methods lack the welldevellinear regression.: the smallest scale distinguishgression is performed properly.: procedures are less sensitive to cal axis onto a variable plotted ~ subsets of extreme values than on the horizontal axis. Resistantfit a variable plotted on the verti. their from a datum to the line.I those of comparable leastsidual is the "vertical" distance ~ squares methods. Re. as well as con. If linear re. However. To fmd the resultant vec• resistant I tor. The if corrupting a datum cannot : resultant vector's tail is the comchange the statistic much. Its head is the image *================ . a re.~ ciated with the leastsquares fitgression is inappropriate. shape that are due to changes negative (if the datum is below : in the positions of just a few the line). Plots of residuals can ~ landmarks. As such. ~ able by a digitising.: oped distributional theory assoditions under which linear re.I • resultant vector ~ the result of combining two vecputational error somewhere.
local maximum. The perpendicular to the curve is a I other two sides are the legs. • rhombus a parallelogram with four equilateral sides I • right triangle I a triangle with a right angle. • ridge curve : The side opposite the right ridge curves are curves on a surface along which the curvature ~ angle is the hypotenuse.146 of the head of one of the vec. Also known as I a vector sum. =================* • right cylinder a cylinder whose direction of sliding is perpendicular to the plane of the base right prism a prism whose direction of sliding is perpendicular to the plane of the base I • I Light Ray :~I""""" • review mandala a circular design arranged in concentric arcs. and size.~ tors after you translate it along. • right angle an angle whose measure is 90 degrees • right cone a cone whose axis is perpendicular to the plane containing its base I I I I • rigid rotation an orthogonal transformation of a real vector space with respect to the Euclidean distance metric. the other vector. the line of the jaw or a motion that preserves shape the rim of an orbit. For instance on I • rigid motion a skull. A principal compo . Such transformations leave distances between points and angles between vectors unchanged.
if X is an estimator of the ~ parameter t.RNG acronym for Random Number Generator.RootMeanSquare Error I (RMSE) : the RMSE of an an estimator ~ of a parameter is the square~ root of the mean squared error . It is rotation to new orthogonal . then RMSE(X) = . A canonical variates : of the elements of the list. The angle of rotation . : . (MSE) of the estimator. ~ compute the rms of a list. To analysis does not. . you . The units of RMSE are the : same as the units of the estima~ tor.II rmserrorofregression I rutatWn 147 *================~ nents analysis represents a rigid ~ of the elements in the list. The RMSE of : an estimator is a measure of the ~ expected error of the estimator. average Y2 : the numbers you get. Mnth_tics======== II . a measure of the average "size" axes. . the nns error of regression is equal to (I r2)!f2xSDY.rotation . This is a process which uses a arithmetic algorithm to generate seyuences of pseudorandom numbers. For regressing Y on X. smaller angle formed by the root of the mean of the squares : lines. in effect.rms error of regression the rms error of regression is the rms of the vertical residuals from the regression line.rootmeansquare (rms) ~ has twice the measure of the the rms of a list is the square. where r is the correlation coefficient between X and Y and SDY is the standard deviation of the values of Y. For: mally. and take I : the squareroot of that average. ~ . In sym: boIs. ~ intersect at the center of the where n is a positive integer. a rotation is an isometry ~ that is the composition of re. a rotation is a turning : of the plane about a point (the ~ center of rotation) by an angle I (the angle of rotation).. I . square all the entries.root of unity I flections through two lines that a solution of the equation xn = I. : rotation. (E( (Xt)2 ) ) !f2.
. The expected value of the sample mean is the population mean. divided by the squareroot of the sample size. such as the percentage of voters registered as Democrats in a simple random sample of voters. {xl. • sa surface area • sample a set of individual units.tUnudsymmetry I samplepercentage II • rotational symmetry a figure has rotational symmetry if it can be rotated (turned) less than 360 degrees about a point so that it appears the same as the original figure. It is a statistic commonly used to estimate the population mean. to be used for a statistical examination of which the findings are intended to be applied to the population. in rerandomisation statistics the process of applying inferences based upon the sample to the population is essentially inforI rna. + xn)jn. the SE of the sample mean is the finitepopulation correction ((Nn)j(Nl»V2 times the SE of the sample mean for sampling with replacement. drawn from some definable population of units. The sample mean is (xl + x2 + . • rusty compass ~ I I a pair of compasses that are fixed open in a given position.xn}. it is essential for such inference that the sample should be representative. sample percentage the percentage of a random sample with a certain properry. The sample mean is a statistic commonly used to estimate the popula I I • I I I I I II ===================Msthematies . Suppose there are n data. and generally a small proportion of the population. For sampling with replacement.. x2. . I • sample mean the arithmetic mean of a random sample from a population. ... the SE of the sample mean is the population standard deviation. For sampling without replacement.=l48==~==~====. • ruled surface a surface formed by mqving a straight line (called the generator). with N the size of the population and !1 the size of the sample.
with N the size of the population and n the size of the sample. • sample size the number of experimental units on which observations are considered. ~ • sample sum . ~ ~ 149 tion percentage. The SE of the sample percentage is often estimated by the bootstrap. Suppose there are n data. with mean M = (xl + x2 + . The SE of the sample percentage for sampling without replacement is the finitepopulation correction ((Nn)/(NI»V2 times the SE of the sample percentage for saQlpling with replacement. . due to the possible multiplying effects of multiple variables and/or repeated measures within the experimental design.. The SE of the sample percentage for sampling with replacement is (p(lp)/n )V2. where p is the population percentage and n is the sample size. xn}. the sum of a random sample : from a population.. It is quite similar to the standard deviation of the sample. Then s = ( ((xl M)2 + (x2 M)2 + .II samplesize I sample sum . . : I : I ~ : I . this may be less than the number of observations in a dataset.. The ex~ pected value of the sample MnthfR'llBtW======== II . : ~ • sample standard deviation the sample standard deviation S is an estimator of the standard deviation of a population based on a random sample from the population. The sample standard deviation is a statistic that measures how "spread out" the sample is around the sample mean. {xl. The expected value of the sample percentage from a simple random sample or a random sample with replacement is the population percentage. S2 (the sample variance) is an unbiased estimator of the square of the SD of the population (the variance of the population). x2... : ~ I : ~ I : ~ I ~ : ~ ~ : ~ ~ : ~ . + (xn M)2)/(nl) )lh The square of the sample standard deviation... but instead of averaging the squared deviations (to get the rms of the deviations of the data from the sample mean) it divides the sum of the squared deviations by (number of data 1) before taking the squareroot. + xn)/n.
I ~ . ..sible' samples. It is an unbiased I estimator of the square of the : population standard deviation.sample survey a survey based on the responses of a sample of individuals. the difference between the estimator and the parameter can be written as the sum of two components: bias and sampling error. The bias is not random. The tool on this page allows you to explore empirically the sampling distribution of the sample mean and the sample percentage of random draws with or without replacement draws from a box of numbered tickets.sampling distribution I I I the sampling distribution of an estimator is the probability distribution of the estimator when it is applied to random samples. with N h . _ lin samp g error in estimating from a random sample. For sampIing with replacement. the SE of the sample sum is the finitepopulation correction «Nn)j(Nl))V2 times the SE of the sample sum for sampIing with replacement. f th I· t e Slze .sample a sample is a collection of units from a population. ~ which is also called the variance I of the population. the SE of the sample sum is the population standard deviation. . The sampling error is random: it comes from "the luck of the draw" in which units happen to be in the sample. : ~ I ~ I ~ : I : I I I . The bias is the average error of the estimator over all pos. . For sampling without replacement. times the squareroot of the sample size. rather than the entire population.sample variance I the sample variance is the square of the sample standard I deviation S.0 e popu atl~n and n the Slze of the samp e. The average of the sampling error over all possible samples (the expected value of the sampling error) . Sampling error is the component of error that varies from sample to sample.150 . I samplesuwey I samplingerror II sum is the sample size times the population mean. It is the chance variation of the estimator.
the ycoordinate of that point would be the weight of the same person . Multivariate statistical findings in the form of coefficient vectors can usually be more easily interpreted if scores are also shown case by case. . The standard error of the estimator is a measure of the typical size of the sampling error. • scale factor the ratio of corresponding lengths in similar figures . A scatterplot is a plot of pairs of measurements on a collection of "individuals" (which need not be people).II sampling unit I s c o r e . I I I I I I I I I I MRthematics======= II . • scatterplot a scatterplot is a way to visualise bivariate data. The fundamental unit of the sample is called the sampling unit. For example. The scatterplot of those data would be 100 points. • scientific notation a notation for expressing very large and very small numbers as a product of a decimal number greater than or equal to one and less than ten and a power of ten. their loadings (correlations with the original variables). The coefficients for the linear combination are usually determined by some matrix computation. suppose we record the heights and weights of a group of 100 people. their scatters. • score a linear combination of an observed set of measured variables. In a scatterplot of height against weight. In a scatterplot of I I I I I weight against height. I I 151 is zero. • scalene triangle a triangle with three sides of different lengths. Each point represents one person's height and weight. the xcoordinates would be the weights and the ycoordinates would be the heights . It need not be a unit of the population. the xcoordinate of each point would be height of one person. or more than one unit at a time (one can sample clusters of units). etc. • sampling unit a sample from a population can be drawn one unit at a time.
• segment of a circle the region between a chord and the included arc. a line that goes through the point of averages. • section (of a solid) An intersection with a plane. If the variable plotted on the horizontal axis is called X and the variable plotted on the vertical axis is called Y. I secular trend a linear association (trend) with time. the set of points consisting of two distinct points and all inbetween them. I I I I I I I I I I I I I • selection bias a systematic tendency for a sampling procedure to include and/or exclude units of a certain type.and the standard error of the sample mean of n random draws with replacement from a box of tickets is • secant a line that intersects a circle in two points. with slope equal to the ratio of the standard deviations of the two plotted variables. the slope of the SD line is the SD of Y. Selection bias is a potential problem whenever a human has latitude in selecting individual II = = = = = = = M 4 t h e m 4 t k s . • se(sample sum) = nl/2 x SD(box). divided by the SD of X. • segment aka line segment. For example.where SD(box) is the standard deviation of the list of the numbers on all the tickets in the box (including repeated values). • se(sample mean) = n V2xSD(box. in a quota sample. • sector of a circle the region between a central angle and the arc it intercepts.152 sd line I selection bias I • II =========* • sd line for a scatterplot. unconscious prejudices or predilections on the part of the interviewer can result in selection bias.
*================ ~ .I stant (known as the magic selves whether they are in the ~ sum).II selfseIeaitm I set 153 ~ .~ in any row or column is a condividuals decide for them. it results in an observational study.I surrounded by the same arsmo "e " volve selfselectIon: tndlVldu. tn. periment. " king on h ~an h alth. a part of itself. ~ same order. Also called a 1Selfselection precludes an ex. _ semtre ar t esse11a0'on 'gu1 ".: reg~ ar PI? ygons hOw" ose f h ffi f I vertIces le on ot er vertIces.selfsimilarity : the sum of a fmite or infinite I the property of a figure that it : sequence is similar to.septagon is selfselection. I r units for the sample. ment group" SelfselectIon" a t esse 11 a t"lon conS1S t'lllg 0 f " " " " IS I qUlte common III studles of i l l all f h human behaviour.: t f' I ( t' J: h 1 I range men 0 po ygons 0 a Is choose lor t emse ves : one or more k' d s )" t h e In III whether or not to smoke. control group or the treat. : without regard to their order. .i uniform tiling. scribed order: a ~. iseries .. ~.sequence from factors that influence ~ a collection of numbers in a preindividuals' decisions to be..semicircle i an arc of a circle whose end: points are the endpoints of a ~ diamete~  .semimagic square i a square array of n numbers . a . ••• p 4 long to the treatment group.selfselection : such that sum of the n numbers selfselection occurs when in. i a set is a collection of things. or approximately ~ . it tends to be eliminated by probability sampling schemes in which the interviewer is told exactly whom to contact (with no room for individual choice).set similar to. " amp1 stud les 0 tee ect "0 : an d"tn W h"lC h every vertex IS e. For ex. one must be : a sevensided polygon I wary of possible confounding : . When there i.
rotation. I I I I • shape the geometric properties of a I configuration of points that . Kendall's shape space. For data that are con.I . more generally. as single I points in a space. Now. mvanant to ch anges m . rotation. without any : nuisance parameters (posi. shape space a space in which the shape of a figure is represented by a single point. scale)..g. I Other sorts of shapes (e. and any of the sets of shape coordinates that arise in geometric morphometrics. Useful shape variables include angles. with a geometry given by Procrustes distance. ratios of distances. • seven bridges of konisberg network a collection of designated points connected by paths. surfaces. coordinates with respect to any basis for the tangent space to Kendall's shape space in the vicinity of a mean form. . we represent the shape of an ob. that does not change under similarity transformations: translations. I those of outlines. ject by a point in a space of ~ shape variables.. . I there is also a representation I of shapes per se. der similarity transforma. It is of 2p4 dimens ions for 2dimensional coordinate data and 3p7 dimens ions for 3dimensional coordinate data. • shape variable any measure of the geometry of a biological form.154 set ofdam ptlints I shape lIariable I • II • set of data points data collected and placed into ordered pairs for the purpose of graphing.. which are' ~ measurements of a geometric object that are unchanged un. or the image of a form. or : functions) correspond to quite I different statistical spaces. and I scale. translation. rotations. shape coordinates in the past. I. tions.: figurations of landmarks. any system of distanceratios and perpendicular projections permitting the exact reconstrUction of a system of landmarks by a rigid trusswork. In morphometries. and changes of geometric scale (enlargements or reductions).~ tion. are . .
null hypothesis when the null : hypothesis is true. : ~ ~ . In multivariate morphometries. a somewhat different use of pure shear is in a transformation of the "shape principal compo~ 155 *================ . significance I level. : I nents" of an allometric analysis of distances to be uncorrelated with withingroup size . ~ • similar polygons . supplies one orthonormal basis for the subspace of uniform shape changes of twodimensional data. : ." geometric morphometricians usually use the word "shear" as an informal synonym for "affme transformation. statistiCal significance.• significance ~ also known as . : • side of a polygon ~ a single segment from the union I that forms a polygon I • sides : (of an angle) The two rays. invariant and alters the other by a translation that is a multiple ofy: for instance. a map taking a square to a parallelogram of unchanged base segment and height. : The significance level of an hy~ pothesis test is the chance that ~ the test erroneously rejects the .themtlties======~ II . • side (of a polygon) A line segment connecting consecutive vertices of a polygon.II shesrlsimilarpolygons • shear in twodimensional problems. I • Sierpinski triangle ~ a type of fractal. Without the adjective "pure. shape aspects of any affine transformation can be diagrammed as a pure shear. hav~ ing a common endpoint. y. that ~ form an angle. same. together with a separate score for change in the horizontal/vertical ratio. lar if their sides are in propor~ tion and all their angles are the . This is a transformation that leaves one Cartesian coordinate." since any 2D uniform transformation can be drawn as one if you wish. what happens when you slide the top of a square sideways without altering its vertical position or the length of the horizontal edges. I : • similar figures ~ two geometric figures are simi. polygons whose corresponding : angles are congruent and whose I MJr. The score for such a translation.
then taking those units flection.: est random numbers to be the tions. Similarities are arbi. : I • Simpson's paradox what is true for the parts is not • simple events a single activity in a probability necessarily true for the whole. and changes of I sample.I tion. ~ ·an experiment that has the I . end.sample of size n from a poputios of distances unchanged. I lation of N > = n units can be The term proper or special simi. same number of outcomes as a • simple polygons convex.one to each unit in the populaformations do not involve re. closed shapes bounded given situation but is easier or by line segments joined end to I more practical to carry out than the given situation.I constructed by assigning a ranlarity group of similarities is dom number between zero and sometimes used when the trans.!!!!:1!!!!!56~~~~~~~~~~:laritytransjimJmtiun lsimulatiun corresponding sides are proportional.draws). the I probability that the sample will r consist of any given subset of n of the N units in the population I is Iren. A simple random nate system that leaves all ra. that is. experiment such as flipping a : • Slmulation I • .that were assigned the n largtrary combinations of transla. rotations. com. scale. . I • II simple random sample a simple random sample of n units from a population is a ran6 I dom sample drawn by a proceL I dure that is equally likely to give every collection of n units 4 from the population. Simple random sampling is sampling at random without replacement (without • similarity transformation I replacing the units between a change of Cartesian coordi.
. orthogonal.size change factor size change magnitude ~ 157 measures of dimension one. . • : . where L is the diagonal matrix of eigenvalues and E the matrix of eigenvectors. • ·thm • · _skew i two lines are skew if they do not ~ intersect and are noncoplanar.singular value decomposition any mxn matrix X may be decomposed into three matrices U. ~ of two sliding logarithmic Interlandmark lengths are size • scales. form (i. i areas are size measures of di: mension two.sine (ofcin acute angle) The ratio of the. e. • .slant height . respectively) in the form: X= UDVt. some measure of a ~ lateral face of a pyramid. D is a diagonal matrix of singular values. an invariant under the group of isometries) that • . : . a distribution that is not sym: metrical. . Mathematics=================== II . V (with dimensions mxm.size measure ~ the height of each triangular in general.Ii sine I slide rule . etc.skewed distribution .: . mxn. where the columns of U are .slide rule scales as a positive power of the : a calculating device consisting geometric scale of the form.skew lines ~ noncoplanar lines that don't • intersect • • o ~ . and the columns of V are orthogonal. .skeleton division ~ a long division in which most i or all of the digits have been : replaced by asterisks to form a : cryptan. D. length of the opposite side to the length of the hypotenuse in any right triangle containing the angle. The singular value decomposition of a variancecovariance matrix S is written as S=ELEt. and nxn.
a volume." The length of the horizontal leg is • space the "run. examples: conic solid. cal leg is the "rise. on the surface of a sphere. treated as if they were the legs between the points. You can calculate the I region of space enclosed by a 3slope m of a line (or line seg. The as a/b dividing the numerator length of the triangle's verti. • slope of a line I • small circle in a coordinate plane.Xl) I • • slope triangle a right triangle used to help fmd the slope of a line or line I segment through two points. • solid The slope of a vertical line is the union of the surface and the undefined. I • slope I sptUe II I I I slopeintercept form the form of a linear equation y = fiX + b where m represents the slope and b represents the yintercept. or on any higherdimensional solid of revolution a solid formed by rotating a twodimensional figure about a line. a collection of obto each quantity depending on : jects or measurements of obthe direction of travel along ~ jects.158 • slope the ratio of the increase in the yvalues to the increase in the xvalues between any two ordered pairs. II ====================Mathematics . points in a plane. I • solidus which are used as the end.cylindric solid.I D figure. .yl) and (x2.the slanted line in a fraction such points of the hypotenuse. ment) through points with co.Yl) I mensional space ( ~ . rectangular solid ordinates (xl.I from the denominator. ." Signs are attached I in statistics.y2) I • solid geometry using the formula m = the study of figures in threedi(Y2 . the the circle formed by the interamount of vertical change : sectiCJn of a sphere and a plane (change in y) for each unit of I that doesn't contain the center horizontal change (change inx).
: the expected squared difference I between the random variable • square a quadrilateral with 4 equal . a variable (a set of data) is said the SquareRoot Law says that: to be in standard units if its the standard error (SE) of the ~ me~n . The SE sides and 4 right angles. That is. linear vector spaces. and di~ viding the results by the stan• standard deviation : dard deviation. The : SE of a random variable X is I . n2. that is. the form of a number expressed : as a sum of products involving perfect square. • square free an integer is said to be square ~ • standard form free if it is not divisible by a .II sphere IsttuuJ. shape spaces. etc. for n> 1. ment of the set and the mean : of the set.ero and its standard sample sum of n random draws ~ dCV1anon IS ~ne. 159 *================ of deviations between each ele.sphericru oigonometty : defmed to be SE(X) = [E( (X the branch of mathematics dealI E(X))2 )] V2. the staning with measurements on the ~ dard error is the squareroot of sphere. I ~ : • standard error ~ the Standard Error of a random . and its expected value. a number of the form n2 • ~ • standard units • squareroot law . a set of dat~ mto standard uruts tickets with numbers on them : by subtracnng the mean from IS I each element of the list. I in repeated experiments.is z. variable is a measure of how far • sphere the locus of pointsin threespace : it is likely to be from its exthat are a fIXed distance froma ~ pected value.srd units generalisation of these intuitive strUctur~s. Examples are: Euclidean spaces.m with replacement from a box of . A random varithe standard deviation of a set ~ able is said to be in standard of numbers is the rms of the set Mathcmatics======= II . You transfor. its scatter given point (called the center). • square number ~ powers of ten. sample spaces. : of a random variable is analo~ gous to the SD of a list.
This situameasurement scales according : tion occurs when attempting to the meaningfulness of arith. standa1"1lise I stmtifted I II units if its expected value is zero and its standard error is one. Sta. The types are similar independent experi: nominal scale versus ordinal ments. a statistic is a random variable.: marked tool used to draw known parameters. The data • Stevens' typology this is widelyobserved scheme : from the various strata are reof distinctions between types of I garded as distinct. and to test hypoth. • standardise to transform into standard units. with each such further where the leading digit (s) are I set distinguished by a level of the stem and the ending single a categorical variable which is dio1ts are arranged in ascend. I experimnad design within a single set (stratum). straight angle an angle whose measure is 180 degrees. You transform a random variable to standard units by subtracting its expected value then dividing by its standard error. I • I I I • • statistic : • straightedge.mental design whereby a scheme of observations is reeses. unmarked a number that can be computed I just how it sounds. forming a line with its sides straightedge a tool used to construct straight lines.I 0distinct from any categorical ing order to the side represent.I this is a feature of an experirameters.• stratified tistics are used to estimate pa.to make inferences based metic which may be performed I upon the results of several upon data values.I variables used to define the ing the leaves. As a func. I scale versus interval scale versus ratio scale. .I straight lines tion of a random sample. I peated entirely using further • stem and leaf plots sets (strata) of experimental a method of displaying data I units. involving no un. an unfrom data.160 .
. rather than from the frame as a whole. it would be advantageous to stratify on geography. sample standard deviation S. For example. If the . :" tic. someI timated standard error of the times the sample is drawn sepa.) Samples drawn in such a way are called stratified samples. Each such subset is called a stratum.. subsets of sampling units are selected separately from different strata. sampling the act of drawing a stratified ~ • studentised score . Estimators based on stratified random samples can have smaller sampling errors than estimators computed from simple random samples of the same size. Student's t curve is used . : Define the random variable T ~ = (M J. called the degrees of freedom. are nearly normally distrib: uted and have population ~ mean is J. and computing the ~ sample mean M and the .. ~ 2. to approximate some prob: ability histograms. then divide each ~ state into urban. statistic. rural areas. if the average variability of the variable of interest within strata is smaller than it is across the entire population..L.. if stratum membership is associated with the variable. and . suburban. Consider a ~ population of numbers that . divided by the esin random sampling. the : probability histogram ofT can I Msth_ti&s======= \I . to determine average home prices in the India. because average home prices ~ • student's t curve : student's t curve is a family of ~ curves indexed by a parameter .II stmriftetlsmnple I stulimt'stcurPe • stratified. We might divide the country : into states.L)/(S/nV2). minus the expected value of • stratum ~ the statistic. then draw random : samples separately from each ~ such division. . that is. rately from different disjoint ~ vary enormously with location. sample in a stratified sample.. sample size n is large. 161 • stratified. or subsets of the population. : which can take the values 1. Consider drawing I a random sample of size n with : replacement from the popula~ tion. the observed value a statissample. (The plural of stratum is strata. .
P(a < T < b) is it came from). Every ele. Student's t 1 nal set need' be in a subset curve with n I degrees of free. Student's t curve can be .not every element of the origiate values of n. h cl th degrees of freedom.162 ~ ~tslmbjea I superimposifitm ..always be identical to the set tion.g.subroutine 1 marks. I geometric relationship to an_ subject. •• Student's T curve with n I : . but ever." e. eac one oser to e desired : value. experimental other figure. ~~~~~~~~. How. Sometimes informally a subset of a given set is a col.I (otherwise. That is. referred to as a "fit" or "fitlection of things that belong: ring. or in other _ subset ~ ways.~ . This page con. a subset would dom gives a better approxima. tains a tool that shows _ superimposition Student's t curve and lets you ~ the transformation of one or find the area under parts of the 1 more figures to achieve some curve. long to the original set.successlve apprOXimatiOn 1 a sequence of approximations.I .. by leastsquares optia previously known algorithm: mization of squared residuals used in another algorithm ~ at all landmarks. where the term known to be nearly normally 1 is in the consequent. The transformaI tions are usually affine transsubject a member of the control group I formations or similarities. II be approximated accurately ~ ment of the subset must beby the normal curve.sufficient condition struct confidence intervals for 1 a version of a conditional that the population mean. from a to· b. or the treattnent group. used to test hypotheses about the population mean and con. a condition distributed. approximately the area under I .1 that implies a preset conclusion.. a resistant fit. when tells you when you can use the the population distribution is term defined. for small and intermedi. : They can be computed by ~ matching two or three land. to the original set.
• symmetric property of congruence • supplementary angles 2 angles whose measures. ures that. the sum of the ar~ larly bisects. for every value of b. For ~ example. a vertical line drawn at x=a. when ~ the property of congruent figadded together.systemRtJ& error 163 ~ gram will be symmetrical about • supplementary two angels are supplementary . an error that affects all the : measurements similarly. the histogram or the probability his to I I symmetry line ~ the line of reflection in a reflec: tionsymmetric figure I • : • system of equations ~ a set of two or more equations. . the boundary of a 3D figure I • symmetry diagonal • surface area : the diagonal that perpendicu(of a solid). A list of numbers has a symmetric distribution if there is a number a such that the fraction of numbers in the list that are greater than or equal to a + b is the same as the fraction of numbers in the list that are less than or equal to ab. ~ • systematic error . if one geometric fig: ure is congruent to a second figgrees ~ ure. equal 180 de.II supplmuntllry 1. • symmetric distribution the probability distribution of a random variable X is symmetric if there is a number a such that the chance that X> =a + b is the same as the chance that X < =ab for every value ofb. In either case. I MnthemRncs============ II . then the second figure is • surface ~ congruent to the first. the other and is a eas of all the surfaces. ~ symmetry line for the kite • symmedian reflection of a median of a triangle about the corresponding I angle bisector. of they add up to 18()<>.. if a ruler is too short.
t i c s .164 systematic random sample I tltm I II =================* everything measured with it will appear to be longer than it really is (ignoring random error) . the eJev.I ten behave essentially as if : a multivariate generalisation of the univariate t 2 statistic. Svstematic random sampIing is better than systematic sampling. test an hypothesis test based on approximating the probability histogram of the test statistic by Student's t curve. • t2 statistic t I • I I • systematic random sample a systematic sample starting I at a random point in the listing of units in the of frame. then the 'P II = = = = = = = M R . and we I want a sample of size N/I0. ignoring random error). For example. if there are N I units ifi the frame. the 21st unit. eLc. or comparing the difference between two mean vectors to a parametric difference (usually the zero vector). but typically not as I good as simple random sam. but they of. I • systematic sample a systematic sample from a I frame of units is one drawn by I listing the units and selecting I every kth element of the list. instead of starting at the first I unit. if the order in which the units appears in the list is haphazard. If the observations are independently multivariate normal. t he square 0 f t h e ratIo 0 f th e group mean difference to the standard error of that difference.I enth unit. Used in the 'Ptest. t tests usually are used to test hypothes. • t 2 test a test due to Hotelling for comparing an observed mean vector to a parametric mean. Systematic samples are a special case of cluster samples. If your watch runs fast.es about the mean of a population when the sample size is intermediate and the distribution of the population is known to be nearly normal. Systematic errors do not tend to average out. t h e m R . every time interval you measure with it will appear to be longer than it really is (again. It is .I piing. Systematic samples are not I random samples. we would take every tenth I unit: the first unit. I I I they were random.
II :able ojJlalues I tangent space
\
I
165
*=================
test may be used to test null hypotheses using the Fdistribution. T2 is also closely related to Mahalanobis D2. • table of values a table of two colLUnns, the first representing values of the independent variable, the second representing the values of the dependent variable.
I I
actual outcome as part of the tail.
tangent (of an acute angle) The ratio of I the length of the opposite side I to the length of the adjacent side in any right triangle containing the angle. I • : . tangent clr~1 es
I •
I circles that are tangent to the : same line at the san1e point. • tail an area at the extreme of a ~ They can be internally tangent randomisation distribution, I or externally tangent. where the degree of extremity is sufficient to be notable judged against some nominal I alpha criterion value.
• tail definition policy this is a defined method for dividing a discrete distribution into a tail area and a body area. the scope for differing policies arises due to the noninfinitesmal amount of probability measure which may be associated with the actual outome value. The conventional policy, based upon considerations of simplicity and of conservatism in terms of alpha, is to include the whole of the weight of outcomes equal to the
)
I
: • tangent line ~ a line that lies in the plane of a ; circle and that intersects the : circle at exactly one point (the ~ point of tangency).
~ • tangent segment ; a line segment that lies on a tan: gent line to a circle, with one ~ endpoint at the point of tanI gency.
I • tangent space : informally, if S is a curving space ~ and P a point in it, the tangent
MRth_ties================== II
166
tangmtitU veIoeity I tensor
II
========*
space to S at P is a linear space I it left the circle along a tangent T having points with the same I line. "names" as the points in S and I • tautology in which the metric on S "in the a sentence that is true because vicinity of P" is very nearly the I of its logical structure. ordinary Euclidean metric on T. One can visualise T as the pro I • tensor jection of S onto a "tangent an example of a tensor in plane" "touching" at P just like morphometrics is the represena map is a projection of the sur tation of a uniform component face of the earth onto flat pa I of shape change as a transforIn geometric I mation matrix. The transformaper. morphometrics, the most rel tion matrix assigns to each vecevant tangent space is a linear I tor in a starting (or average) vector space that is tangent to form a vector in a second form. Kendall's shape space at a point A rigorous, general definition corresponding to the shape of ~ of a tensor would be beyond a reference configuration (usu . the scope of this glossary, but a intUltIve ally taken as the mean of a reasonably sample of shapes). If variation I characterisation comes from in shape is small then Euclid I Misner, Thorne, and Wheeler, ean distances in the tangent Gravitation (Freeman, 1973): a space can be used to approxi I tensor is a "geometric machine" mate Procrustes distances in I that is fed one or more vectors Kendall's shape space. Since the in an arbitrary Cartesian coortangent space is linear, it is pos dinate system and that produces sible to apply conventional sta I scalar values (ordinary decimal tistical methods to study varia numbers) that are independent tion in shape. of that coordinate system. In I morphometrics, these "num" • tangential velocity bers" will be ordinary geomet(0f an object moving in a circle) I ric entities like lengths, areas, The speed of the moving object I in the direction it would take if : or angles: anything that doesn't I change when the coordinate system changes. For the represen
II=======MII~
II terminal side I tetrahedron
167
tation of a uniform component ~.  tessellation as a transformation matrix, the I an arrangement of shapes "scalars" of the Misner.:rhorne : (called tiles) that completely Wheeler metaphor arc the ~ LOvers a plane without overlaps lengths of the resulting vectors I or gaps. and the angles among them. A different tensor representing the same uniform transforma I tion is the relative metric tensor, which you probably know as the ellipse of principal axes and I principal strains. This tensor produces the necessary numerical invariants (distances in the ~ second form as a function of ; . . .  test statistic coordmates on the first form) : . . d h th directly. Other tensors include ~ a sta~ch use hto . test ypo ~ the metric tensor of a curving ; eses. dYPOb t des1s'dt~st can di . constructe y eCl 109 to re, hi sun.ace W ch expresses stance·. th ull h h' h th on the surface as a function of ~ Ject e n ypot eSls.w. e~ . e . c. the parameters 10 wh'1Ch sun ace I. value of the test statistic IS 10 . . d d h some range or collection of pomts are expresse an t e T . h cUrPature tensor of the same sur I ran~es. J.~ ~et a test WIt a c. hi h· . lace W ch expresses te way 10 I speCified slgmficance level, the · C. chance when the null hvpothes1s lauS . . . w h1C h the surlace "C._II awa y" fi' I I IS true that the test statistic falls r~m itS tangent p ane at any ~ in the range where the hvpothpomt. . . : eS1S wo uld be rqected ' must be  terminal side ~ at most the specified signifithe side that the measurement ; cance level. The Z statistic is a of an angle ends at : common test statistic.
I
*=================
 tesselate : the ability of a regIOn to ~ tessalate ; :
I
 tetrahedron a polyhedron with four faces. The regular tetrahedron is one of the Platonic solids.
MR.,hmuJhu======= II
• thinplate spline in continuum mechanics. uniform plate constrained only by displacements at a set of discrete pointscan be solved algebraically by a simple matrix inversion.) One particular version of this problem an infinite. space) II =======MRthmulriu . then use valid forms of reasoning to arrive directly at a conclusion. In morphometrics. In that form. I I I I I • the five platonic solids the five regular polyhedrons: regular tetrahedron. the same interpolation (applied once for each Cartesian coordinate) provides a unique solution to the construction of D'Arcy Thompsontype deformation grids for data in the form of two landmark configurations. regular icosahedron. I • I I I I • threedimensional having length. the technique is a convenient general approach to the problem of surface interpolation for computer graphics and computeraided design. (The extent of bending is taken as so small I I I I I I I that elastic energy stretches and shrinks in the plane of the original plate can be neglected. and regular dodecahedron. and/or previously proved theorems. width. • theorem important mathematical statements which can be proven by postulates.168 • tetromino a foursquare polyomino. regular octahedron. and thickness (i.e. a thinplate spline models the form taken by a metal plate that is constrained at some combination of points and lines and otherwise free to adopt the form that minimises bending energy. definitions. regular hexahedron. three types of proofs direct proof a proof in which you state premises..
whether or not data are rep [resnted by ranks.5. the ranks assigned to a set of 6 data. the pinball machine too hard.6.5.3.5.2. I all a l2 an !lIZa]1 a]Z I a l3 a lZ all al3 I al2 aJ3 a l2 all . MRtm#Cs==================== II . I : .Toeplitz matrix a matrix in which all the ele~ ments are the same along any .3. . . tied values lead to tied ranks. a "doughnut".tied ranks in a non parametric test involving ranked data. for any test statistic the occurrence of tied values will increase the extent to which a randomisation distribution will be a discrete distribution rather than a continuous distribution.traee ~ the trace of a matrix is the sum . _ tied values where data are represented by ranks.6 or 1. it is generally agreed that this should be the average of the ranks which would have been assigned if the values had been discernably unequal. if two data have tied values then they will deserve to receive the same rank value. The possibility of tied ranks leads to elaborations in the otherwisestandard tasks of computing or tabulating randomisation distributions where data are replaced by ranks. "Tradi~ tional morphometries" differs . diagonal that slopes from north: west to southeast.torus a 3D figure formed by rolling I a rectangle into a cylirtder and I bending the cylinder until its bases meet. I : . with ties present might emerge as sets such as : 1. .5. Thus.tilt : from the geometric the measure of an angle as ~ morphometries discussed here compared to a horizontal line.traditional morphometries ~ application of multivariate I statistical methods to arbi: trary collections of size or ~ shape variables such as · disI tances and angles.II tied rilniu I trlUlitiolUJl mM1 hometrics • =========1=6=9 ~ what happens when you kick . .3.3.3. of the terms along the princi: pal diagonal.
transformed value of the 1ample of an affine transforma.~ and percentiles are reversed: enheit. are still transformed by into the corresponding list of : the same rule.I lute value of the number you thing (y = ax). In defined to record biologically that case. or variables into I acdy the same formula. median. it's the familtion on the SD. This is an ex. . quartiles. the mean. tiplicative constants have a • transcendental number simple effect on the mean.I formations are used to put tances or measurements are I variables in standard units .by the SD. you multiply each ele. peratures in degrees Celsius I mode. but the geometrical I the SD. so it is an affine erence to Procrustes distance transformation. transformed using exother lists. I the old value times the absoyou only multiply by some. but quartiles temperatures in degrees Fahr. When other variables. I median. you subtract the meaningful aspects of the or.multiplied the first list by: . the multiplicative constant is to transform a list of tem. The efb is the general affine trans.I the qth quantile of the transment by 9/5. mode. range. other percentiles: the new • transformation value of any of these is the old transformations turn lists into lone. and a number that is not algebraic. IQR. and add 32 to formed distribution is the each product.I feet of an affine transformaformation of X. divided morphometrics makes no ref.=17=O=========~talnumber I transformation II in that even though the dis. Affine transor any other aspect of ~ formations with positive mulKendall's shape space. Traditional I negative of the mean.I qth quantile of the original tion: multiply by something distribution (ignoring the efand add something (y = ax + fect of data spacing). For example.I negative. This is equivalent to relationships between these multiplying by the reciprocal measurements are not taken of the SD and adding the into account. is to make the new value In a linear transformation. and iar equation of a straight line). Affine trans.I mean and divide the results by ganism.
congruent to a third. . same length (the distance of the : translation) . The par: allel sides are called bases.transversible a network in which all arcs can 2 ~ be traced without going over what you added does not affect lone more than once them.transitive property of .translation vector ~ see translation ~ . . . A ~ pair of angles that have a base . .transversal . if one geo. Establation vector) between each : lishing whether the treatment point and its image are all parallel (determining the direction . I ~ : . then the first and third objects are congruent to each other. and have the . also Sk where the transformation S that maps (x.translation : . I 'd : .trapezium : a quadrilateral in which no sides I : are parallel. y) onto (kx. the property that.II tmnsformationnotation I treatment~==========17=1 of the translation).~ the variable of interest.treatment effect an isometry under which the I the effect of the treatment on vectors (any of which is a trans.trapeZOl I a quadrilateral with exactly one ~ pair of parallel sides. which stands for the transformation of P. . as a common side are called a congruence : pair of base angles.transformation notation t(P). ky) and k is the magnitude of that transformation I .I metric object is congruent to a second object. which in turn is . a line that intersects 2 others .
• tree a tree is a graph with the property that there is a unique path from any vertex to any other vertex traveling along the edges. triangular numbers numbers of dots that can be put into triangular arrangements. . as opposed to those in the control group. periodic curve A curve that repeats in a regular pattern. • trigonometry the study of the relationships between the measures of sides and angles of triangles. 0) and radius 1 unit. • triangle a polygon with three sides.ries . I • I I I • triangulate to divide a polygon into triangles • tridecagon a 13sided polygon • trigon a threesided polygon. • treatment group the individuals who receive the treatment. • triangle inequality the property that states that the length of any side of a triangle is less than the sum of the lengths of the other two sides. trigonometric ratios and the unit circle unit circle a circle on the coordinate plane with center (0. who do not. : ~ . equivalently. • tree diagram a concept map in the form of tl1e branches of a tree. trinomial an algebraic expression consisting of 3 terms. I I I • I II =======MlJtbem. At issue is whether the treatment has an effect on the outcome or variable of interest. sums of consecutive positive integers beginning with 1. period The horizontal distance between corresponding points on adjacent cycles of a periodic curve. You can use it to show the relationships among members of a family of concepts. I I I • I .172 matmmtgroup I trinomud I II =================* has an effect is the point of an experiment. • treatment the substance or procedure studied in an experiment or observational study.
esis test of the null hypothesis : that the value of a parameter. I designed to have power against : the alternative hypothesis that ~ either /L < /LO or /L > /LO (the I alternative hypothesis contains ~ values on both sides of the null : value). a signifII cance level 5%. ~ : _ twocolumn proof a form of proof in which each statement in the argument is written in the left column. twosided z test ~ of the null hypothesis that the : mean of a population equals ~ zero against the alternative that . For example..twosided hypothesis test ~ c. . after forms are rescaled and . 4. but no thickness . /LO. I : . An hypoth.. onesided test.tromino a threesquare polyomino.twopoint perspective a method of perspective drawing that uses two vanishing points. 0) in a Cartesian coordinate system. repositloned so that landmark 1 is fixed at (0. truncated pyramid part of a pyramid remaining I after truncating the vertex with a plane parallel to the ~ base. For example. ~ . and the reason for each statement is written directly across from it in the right column. _ twocolumn proof a form of proof in which each statement in the argument is written in the left column. consisting (for twodimensional data) of the coordinates of landmarks 3.twin primes I two prime numbers that differ : by 2. 11 and 13 I are twin primes. and the reason for each statement is written directly across from it in the right column. '. originally Francis Galton's.II tromino I twosided hypothesis test ~ 173 *================ .0) and landmark 2 is fixed at (1. Also referred to as Bookstein coordinates or Bookstein's shape coordinates.twodimensional having both width and length. is equal to a null value.. _ twopoint shape coordinates a conveniel1t system of shape coordinates. . rediscovered by Bookstein.f. : I : . it is greater than zero would : reject the null hypothesis for ~ values of 1\ . ~ /L.
• type 2 landmark a mathematical point whose claimed homology from case to case is supported only by geometric..be kept in mind in the course gest evidence.I resent one scheme of alternatives covering the whole of the • type 3 landmark the data represented. I morphometries permits them to be treated as landmark • type 1 landmark a mathematical point whose I points in some analyses. q.I a landmark having at least one umns represent a further I deficient coordinate.174 =================* (sample mean) I Izl=> true. the null hypothesis is rejected erroneously when it is in fact I • unbiased not biased. . I I ~ 1. the null hypothesis is not re: jected when it is in fact false.96. not histological.I interpretation. and the entries in the I concavity. A 1YPe 2 error occurs if .I gion of the form. . Type 3 landmarks twoway table are the counts of : characterise more than one renumbers of observations con. such as a local of any geometric or biological pattern of juxtaposition of tis. The multiforming to the respective cells I variate machinery of geometric of the twoway classification. having zero bias. A Type 1 error occurs when mate: . it is intended to estiing.I value. evidence: for instance. either end of a longest the whole of the data repre. but the claimed homology from case to I deficiency they embody must case is supported by the stron. such that the rows rep. the col. that has expected value the parametric • type land type 2 errors these refer to hypothesis test. ~ • unbiased estimator as its an estimator.I diameter. for inscheme of alternatives covering stance. or the bottom of a sented. sue types or a small patch of : some unusual histology.SE(sample mean) I • twoway table a representation of suitable data in a table organised as rows and columns. the sharpest curvature of a tooth.
Once a metric is supplied . closest to a particular target : form. the undefined terms are ~ point.uniform shape component ~ that part of the difference in I shape between a set of con~ figurations that can be mod: eled by an affine transforma~ tion. the uni. terms : that are assumed. In our (Euclidean) sysI tern. and space. they are uncorrelated. that uniform _ uncountable a set is uncountable if it is not : transformation is computed ~ by a formula based in countable. Together ~ with the partial warps. line.1~7~5 _ uncontrolled experiment an experiment in which there is no control group. for shape space one can ascer: tain which such transforma~ tion takes a reference form . ing is derived only from the : postulates or axioms that use ~ them. form component defined in : this way supplies an orthonor~ mal basis for all of shape space . _ uncorrelated a set of bivariate data is uncorrelated if its correlation coefficient is zero. : shape coordinates.e. the uniform ~ shape component may also be MIIthmuJtics========== II .II uncontrolled experiment I uniformS.undefined term . If two random variables are independent. ~ . but the outcome of the treated group is not compared with the outcome of a control group that does not receive treatment.. 1\:vo random variables are uncorrelated if the expected value of their product equals the product of their expected values. in the vicinity of a mean form.. : In this setting.""cmn='P""onetJ=""t======". "and whose mean. and assigned ~ no properties. i. (The converse is not true in general. in a deductive system. shape space). in which the method of comparison is not used: the experimenter decides who gets the treatment. I Procrustes residuals or by an_undecagon ~ other based in twopoint an elevensided polygon.) : . plane. For the Procrustes met~ ric (the geometry of Kendall's .
.unilateral surface natural number as a denomina"d a surface with only one S1 e. the uni.B". B. The lar specimen) into the plane I union of the events A and B is (or hyperplane for data of di. C. or is the set of objects contained between a mean and a particu.I by at least one of the sets.unit square a square matrix is unimodular I a unit square is a square of side if its determinant is 1. or I a divisor d of c is called unitary both. I with one as a numerator and a .UB". I a tmit might be "meters")" _ unimodular . I dius 1.76===========* unilateralsuiftue I unitarydi'Pisor \I interpreted as the projection ~.f. I tor. through that mean form and I _ unit analysis all nearby forms related to it I the process of using conversion by affine transformations. _ unit of attribute _ unimodal I the unit chosen depends upon a finite sequence is unimodal if : the attribute being measured it first increases and then de. _ union of two sets a and b _ unitary divisor the set of elements in A.unit fraction tesian axes.. fixed.I .. form component is parameterised not by a vector.~ (e. II ===================MJJth_tics . such as a Moebius strip.unit cube dilations with respect to a I a cube with edge length 1.I measure or rate to another. intersection. for the attribute "length".. length 1. "A or B".c/d) = 1. creases. For factors to change from one descriptive purposes. in terms of sets of shears and .""1.unit circle like the partial warps. but by I a unit circle is a circle with raa representation as a tensor.g. orthogonal set of Car. written AUB I if gcd(d.denoted "A&plus.I the union of two or more sets tween two group means.. mension greater than two) and '1\.union of a shape difference (be. .
everything that is outside the sets . mean.f. an argument that reaches its : conclusion through accepted ~ forms of reasoning. . bivariate. population variance The vari: ance of a list is the square of the ~ standard deviation of the list. and SD.177 *==~~~~~~~ . The vector's direction is indicated from the tail to the head. : _ Venn diagram ~ a pictorial way of showing the I relations among sets or : . . I _ vanishing line : the horiwn. and its expected value: Var(X) : = E( (X E(X) )2). median. that is.valid reasoning .upper bound any number above which a function value may approach but not I : pass. C. ~ . The variance of a random ~ variable X. ~ . in a drawing it is ~ at the height of viewer's eye ~ _ vanishing point a point toward which lines in a perspective drawing converge if they represent parallel lines that recede from the foreground to the background. square of the standard error : (SE) of the variable. _ universal statement a conditional that uses the words 'all' or 'everything' _ universe in a Venn diagram. The variance ~ of a random variable is the . usually represented by an arrow with a head and tail. Variance. quantiles. I . percentiles.vector a quantity that has both magnitude and direction.velocity ~ the rate of change of position. is the ex: pected value of the squared difI ference between the variable .univariate having or having to do with a single variable. The vector's magnitude is the length of the line segment. IQR. the average of the : squares of the deviations of the ~ numbers in the list from their I mean. Var(X). _ variable i a numerical value or a charac: teristic that can differ from in~ dividual to individual. Some univariate techniques and statistics include the histogram.unity one .
The part of the rect. I • vertex (of a polygon) The point of intersection of three or more edges. I • I I I II =======M#them4ties .I • vertex of a conic solid angle included in one or more the point that marks the thinof the regions corresponds to I nest part of a conic solid the union of the sets. planes. I • vertical line a line that goes straight up and down. the sets are I cal name. polygon an event by the area of the • vertex of an angle event. The universal set or I • vertex arrangement outcome space is usually I a notation that uses positive drawn as a rectangle. vertices plural form of vertex. • vertex angles (of a kite) The angles between the pairs of congruent sides. If the regions tion. sets are integers and other symbols to regions within the rectangle. "corner" point of any geometric figure bounded by lines. This page contains a tool that illus. I describe the arrangement of The overlap of the regions I regular polygons about verticorresponds to the intersec.ces of a semiregular tessellation of the sets. the tool ~ an endpoint of a segment in a represents the probability of . Also called the numerido not overlap. disjoint. or lines and planes. and whose slope is defined as infinite or undefined .178 events.• vertex of a polygon trates Venn diagrams. the common endpoint of the I two rays I • vertical angles nonadjacent angles formed by the intersection of two lines . the point of intersection of the rays of an angle.
• vigesimal related to intervals of 20. the ranks are ~.. for a sample of shapes. although the variants I each correspond to valid well: sized exact tests. which need ~ be only ordinalscale. the test statistic is derived ~ from the set of differences be. : thus or otherwise. For example. vational unit. • Wilcoxon test : named after the statistician F. Wilcoxon's ~ original procedure commences I by discarding entirely the obser~ vations from any experimental : units for which the data values ~ are equal at each level of the . the pur: pose is to measure shift in scale I location between the two levels ~ of the repeated measure distinc: tion. II . the next step ~ is ranking the differences.11l'iew I WU~ test • view ~ • whole number a drawing of a side of an object . ~ Wilcoxon .. tween the two levels of the re: peated measure distinction one ~ difference score for each obser. like them.. • vulgar fraction a common fraction. wff • a wellformed formula. This test applies to I an experimental design involv~ ing ~o repeated measure ob: servanons on a common set of I experimental units. • vinculum the horizontal bar in a fraction separating the numerator from the denominator. viding a rank for each retained : experimental unit. together with the uniform component.. The weight ~atrix is com~uted as a ~otanon of the ProcrustesresIdual shape coordinates. according to the absolute values Mtldlmufriu======.. • weight matrix w matrix :fhe matrix of partial warp scores.. repeated measure comparison. they are a set of shape coordinates for which the sum of squared differences is the squared Procrustes distance between any two specimens. the procedure is : somewhat variable between ~ authors... 179 .. a is less than or equal to b.. pro. • volume the amount of space a 3D object can hold. a natural number. • weak inequality an inequality that permits the equality case.
(pXOR q) is logically equivalent to «p OR q) AND NOT (pANDq». where w = work.180 wintling number I xpentoltlitllJ I II of the differences. The ranks are summed separately into two or three categories: negative differences. given by the formula w = fd. plus an adjustment for the middle (zerodifference) category. • witch of agnesi a curve whose equation xl y=4a 2 (2ay). If p and q are two propositions. the test statistic is the smaller of the outer categories. I • I I IS I • work the result of force applied over some distance. and so on until the data is adequately fit. zero differences (if any). • XOR. • winding number the number of times a closed curve in the plane passes around a given point in the counterclockwise direction. positive differences. and then . I proceeds to fmd additional factors to fit to the residuals. to Sewall Wright. exclusive disjunction XOR is an operation on two logical propositions. due I letter X. (p XOR q) is a proposition that is true if either p is true or if q is true. f = force. in that one fits a simple one factor model iteratively to I maximally explain the correlations among variables. I I I • xintercept the point at which a line crosses the xaxis. xaxis the horiwntal axis in the plane. and d = distance. I • xpentomino • Wright factor analysis a pentomino in the shape of the a version of factor analysis. in which a I path model is used to describe the relation between the mea. ~ sured variables and the factors I of interest. . It is usually exploratory. • x roman numeral for 10. but not both. : ~ .
• yintercept ~ • zero element the point at which a line crosses .• zerodi menSlonaI ~ having no dimension.181 *==~~========~ • yard ~ nal)jSE(original). . • yaxis the vertical axis in the plane. both the initial a measure of time equal to the . • zero divisors earth about the sun. I I . the null hypothesis by the nor: mal curve. I • I • z statistic a Z statistic is a test statistic whose distribution under the null hypothesis has expected value zero and can be approximated well by the normal curve. a point zintercept ~ the point at which a line crosses : the zaxis. an hypothesis test based on ap: proximating the probability his~ togram of the Z statistic under . . .0 . Approxi. I . The Z statistic is related to the original statistic by Z = (original expected value of origi : • zone ~ the portion of a sphere between . • zero angle : an angle whose measure is O.: nonzero elements of a ring mately equal to 365 days. Usually. • year ~ In a zero angle. I • ztest . the element 0 is a zero element the yaxis. : ofagroupifa+O=aandO+a=a : for all elements a. and terminal sides are the same. period of one revolution of the . ~ whose product is O. a measure of length equal to 3 ~ • zero feet. two parallel planes. Z statistics are constructed by standardising some other statistic. • zscore : the observed value of the Z staI risric.