Mathematical Formulae
Measure of Central Value
RELATIONSHIP AMONG THE AVERAGES
A.M (Arithmetic Mean) G.M. (Geometric Mean) H.M. (Harmonic Mean)
Where:
A.M. (Arithmetic Mean) =
u+b
2
G.M. (Geometric Mean) = Vob
H.M. (Harmonic Mean) =
2
1
c
+
1
b
=
2ub
u+b
LIST OF FORMULAE
1. Mean
(In Individual Series)
(Direct Method)
A.M. (Arithmetic Mean) ( X
) =
_X
N
(Shortcut Method)
A.M. (Arithmetic Mean) ( X
) = A +
_d
N
(In Discrete Series)
(Direct Method)
A.M. (Arithmetic Mean) ( X
) =
_]X
N
(Shortcut Method)
A.M. (Arithmetic Mean) ( X
) = A +
_]d
N
P a g e  19
Mathematical Formulae
Standard Error of Estimate for X on Y
S
xy
= _
_(XX
c
)
2
N
; i. c .. . _
Unexpaned Faratun
N
;
S
xy
= o
y
V1  r
2
;
S
xy
= _
_X
2
a_Xh_XY
N
Where:
S
xy
= Ibc StonJorJ Error o Rcgrcssion o :olucs rom X
c
X
c
= Ibc :oluc o X wbicb orc tokcn out rom putting tbc :oluc o in rcgrcssion
cquotion o X on .
o
y
= Ibc StonJorJ Jc:iotion o scrics "Y"
r = Ibc corrclotion cocicicnt
N = Ibc totol numbcrs o obscr:otion
X
2
= Ibc sum o tbc squorc o scrics "X"
_X = Ibc sum o tbc scrics "X"
_XY = Ibc sum o tbc proJuct o tbc scrics "X" onJ "Y"
_(X X
c
)
2
= Ibc Sum o tbc squorc o tbc Jicrcncc bctwccn X  X
c
a, h = "a" onJ "h" orc tbc constonts wbosc :oluc orc tokcn out by sol:ing rcgrcssion
cquotions o "Y un X" onJ "X un Y" simultoncously .
P a g e  2
Mathematical Formulae
(In Continuous Series)
(Direct Method)
A.M. (Arithmetic Mean) ( X
) =
_]m
N
(Shortcut Method)
A.M. (Arithmetic Mean) ( X
) = A +
_]d
N
i
2. Median
In Individual and Discrete Series Size of
N+1
2
th item & in Continuous Series Size of
N
2
th item
Median (Med) = L +
N
2
 c.].
]
i
3. Quartiles (in continuous series)
Q1 =
N
4
th item of the Series
Q1 (Quartile 1) or Upper Quartile = L +
N
4
 c.].
]
i
Q3 =
3N
4
th item of the Series
Q3 (Quartile 3) or Lower Quartile = L +
3N
4
 c.].
]
i
4. Deciles (in continuous series)
D4 =
4N
10
th item of the Series
D4 (Deciles 4) = L +
4N
10
 c.].
]
i
P a g e  3
Mathematical Formulae
5. Percentiles (in continuous series)
P60 =
60N
100
th item of the Series
P60 (Percentiles 60) = L +
60N
100
 c.].
]
i
6. Mode
Steps:
1. The value that occurs largest number of times.
2. Either by inspecting or by grouping method and analysis table determining that value around
which most of the frequencies are it lies.
3. Then apply the formula:
Mode (Mo) = L +
z
1
z
1
+ z
2
i Where: z
1
= f1  fuanu z
2
= f1  f2 (ignoiing signs)
&
Mode (Mo) = L +
]
1
 ]
0
2]
1
 ]
0
 ]
2
i
Fmprca Mude = 3 Medan  2 Mean
7. Geometric Mean
(Individual Series)
G.M. (Geometric Mean) = Antilog [
_Iog X
N
(Discrete Series)
G.M. (Geometric Mean) = Antilog [
_] Iog X
N
(Continuous Series)
G.M. (Geometric Mean) = Antilog [
_] Iog m
N
P a g e  18
Mathematical Formulae
Calculation of Correlation Coefficient in Regression Analysis
Correlation Coefficient of X on Y:
r (curreatun cueJJcent uJ X and Y) = _b
x
b
x
Where:
h
yx
=
_d
x
d
y
 _d
x
_d
y
_d
x
2
(_d
x
)
2
= r
o
y
o
x
; h
yx
is tbc rcgrcssion cocicicnt o on X
h
xy
=
_d
x
d
y
 _d
x
_d
y
_d
y
2
(_d
y
)
2
= r
o
x
o
y
; h
xy
is tbc rcgrcssion cocicicnt o X on
Calculation of Standard Error of Estimate
Standard Error of Estimate for Y on X
S
yx
= _
_(YY
c
)
2
N
; i. c .. . _
Unexpaned Faratun
N
;
S
yx
= o
x
V1  r
2
; S
yx
= _
_Y
2
a_Yh_XY
N
Where:
S
yx
= Ibc StonJorJ Error o Rcgrcssion o :olucs rom
c
Y
c
= Ibc :oluc o wbicb orc tokcn out rom putting tbc :oluc o X in rcgrcssion
cquotion o on X.
o
x
= Ibc StonJorJ Jc:iotion o scrics "X"
r = Ibc corrclotion cocicicnt
N = Ibc totol numbcrs o obscr:otion
Y
2
= Ibc sum o tbc squorc o scrics "Y"
_Y = Ibc sum o tbc scrics "Y"
_XY = Ibc sum o tbc proJuct o tbc scrics "X" onJ "Y"
_(Y  Y
c
)
2
= Sum o tbc squorc o tbc Jicrcncc bctwccn 
c
a, h = o onJ b orc tbc constonts wbosc :oluc orc tokcn out by sol:ing tbc rcgrcssion
cquotions o on X onJ X on simultoncously .
P a g e  17
Mathematical Formulae
Deviations Taken from Assumed Mean
Regression Equation of X on Y
The regression Equation of X on Y is expressed as follows:
X  X
= r
o
x
o
y
(Y  Y
)
Where:
_X
= r
o
y
o
x
(X  X
)
Where:
_X
w
) =
_wX
_w
Weighted Geometric Mean (G.M.
w
) = Antilog [
_wIog X
_w
Weighted Harmonic Mean (H.M.
w
) =
_w
[
1
c
w
1
+
1
b
w
2
10. Combined Mean
Combined Arithmetic Mean (X
123
) =
N
1
X
1
+ N
2
X
2
+ N
3
X
3
N
1
+ N
2
+ N
3
Combined Geometric Mean (G.M.
12
) = Antilog [
N
1
IoguM
1
+ N
2
IoguM
2
N
1
+ N
2
P a g e  5
Mathematical Formulae
Measure of Dispersion
Range
Range = L  S Where: L = Large Value & S = Small Value
Coefficient of Range
Range =
L  S
L + S
Interquartile Range
Interquartile Range =
3

1
Where: Q3 = Third Quartile / Lower Quartile
Q1 = First Quartile / Upper Quartile
Quartile Deviation
Quartile Deviation =
3

1
2
Coefficient of Quartile Deviation
Coefficient of Quartile Deviation =
3

1
3
+
1
Percentile Range
Percentile Range = P
90
P
10
Where: P90 = 90
th
Percentile
P10 = 10
th
Percentile
Semi Percentile Range
Percentile Range =
P
90
P
10
2
Mean Deviation
(Individual Observations)
Mean Deviation (M.D.) =
_
N
Where: D = (X A) (Ignoring Signs )
P a g e  16
Mathematical Formulae
Deviations Taken from Arithmetic of X and Y
Regression Equation of X on Y
The regression Equation of X on Y is expressed as follows:
X  X
= r
o
x
o
y
(Y  Y
)
Where:
_X
) ; y = (Y  Y
)
Regression Equation of Y on X
The regression Equation of Y on X is expressed as follows:
Y  Y
= r
o
y
o
x
(X  X
)
Where:
_X
) ; y = (Y  Y
)
P a g e  15
Mathematical Formulae
Regression Analysis
Regression Equation of Y on X
The regression Equation of Y on X is expressed as follows:
Y
c
= a + hX
For determine the value of a! and b! the following two normal equations are to be solved:
Y = Na + hX
XY = aX + hX
2
Where:
_XY = Iotol o tbc proJuct o Scrics X onJ Scrics
_X = Iotol o tbc Scrics X
_Y = Iotol o tbc Scrics
_X
2
= Iotol o tbc Squorc o Scrics X
a, h = a onJ h orc tbc constonts
Regression Equation of X on Y
The regression Equation of X on Y is expressed as follows:
X
c
= a + hY
For determine the value of a! and b! the following two normal equations are to be solved:
X = Na + hY
XY = aY + hY
2
Where:
_XY = Iotol o tbc proJuct o Scrics X onJ Scrics
_X = Iotol o tbc Scrics X
_Y = Iotol o tbc Scrics
_Y
2
= Iotol o tbc Squorc o Scrics
a, h = a onJ h orc tbc constonts
P a g e  6
Mathematical Formulae
(Discrete Series)
Mean Deviation (M.D.) =
_]
N
Mean Deviation (M.D.) =
_]
N
i
Calculation of Standard Deviation
(Individual Observations)
Standard Deviation ( o ) = _
_x
2
N
Where: x = (X X
)
(Deviation taken from Assumed Mean)
Standard Deviation ( o ) =
_
_d
2
N
 [
_d
N
2
Where: d = (X A); = rcqucncy
Calculation of Standard Deviation
(Discrete Series)
Actual Mean Method
Standard Deviation ( o ) = _
_]x
2
N
Assumed Mean Method
Standard Deviation ( o ) =
_
_]d
2
N
 [
_]d
N
2
Step Deviation Method
Standard Deviation ( o ) =
_
_]d
2
N
 [
_]d
N
2
i
P a g e  7
Mathematical Formulae
Combined Standard Deviation
(When two groups are given)
Standard Deviation ( o
12
) = _
N
1
c
1
2
+ N
2
c
2
2
+ N
1
d
1
2
+ N
2
d
2
2
N
1
+ N
2
Where:
o
1
2
= Square of the Standard Deviation of Group One
o
2
2
= Square of the Standard Deviation of Group Two
d
1
= X
1  X
2  X
1  X
12S (Mean of the first group subtracted by Mean of the combined group)
d
2
= X
2  X
12S (Mean of the second group subtracted by Mean of the combined group)
d
3
= X
S  X
12S (Mean of the third group subtracted by Mean of the combined group)
N
1
= Total number of observation in group one
N
2
= Total number of observation in group two
N
3
= Total number of observation in group three
P a g e  14
Mathematical Formulae
Probable Error
Probable Error (P.E.r) = 0.6745
1
2
VN
Where:
r = Cocicicnt o Corrclotion
N = Iotol Numbcr o 0bscr:otions
Rank Correlation Coefficient (Spearmans Correlation Coefficient)
Rank Correlation Coefficient (R) = 1 
6_
2
_N
S
N_
Where:
= icrcncc bctwccn tbc Ronk A onJ Ronk B
N = Iotol Numbcr o 0bscr:otions
When Ranks are Same / Equal
Rank Correlation Coefficient (R) = 1 
6__
2
+
1
12
_m
S
m] +
1
12
_m
S
m] +_
_N
S
N_
Where:
m = m is tbc somc ronk in tbc scrics or cxomplc i ony ronk orc somc or two
obscr:otion. wc writc it os (2
3
2)onJ i it is somc in tbrcc obscr:otions
tbcn wc writc it os (S
3
S).
N = Iotol Numbcr o 0bscr:otions
P a g e  13
Mathematical Formulae
=
N_X_X_
N
2
N
2
N_X
2
(_X)
2
N_
2
(_)
2
=
N_X_X_
N_X
2
(_X)
2
N_
2
(_)
2
Where:
_X = Iotol o X Scrics
_ = Iotol o Scrics
_X = Iotol o tbc ProJuct o Scrics X onJ Scrics
_X
2
= Iotol o tbc Squorc o X Scrics
_
2
= Iotol o tbc Squorc o Scrics
N = Iotol o tbc 0bscr:otions
r = Cocicicnt o Corrclotion
When Deviation are taken from Assumed Mean
Coefficient of Correlation (r) =
N_d
x
d
j
_d
x
_d
j
_N_d
x
2
(_d
x
)
2
_
N_d
j
2
(_d
j
)
2
Where:
_J
x
= Iotol o tbc Jc:iotion o x scrics tokcn rom AssumcJ Hcon
_J
2
= Iotol o tbc Squorc o tbc Jc:iotion o Scrics tokcn rom AssumcJ Hcon
N = Iotol o tbc 0bscr:otions
r = Cocicicnt o Corrclotion
P a g e  8
Mathematical Formulae
Standard Deviation of N! Natural Numbers
Standard Deviation ( o ) = _
1
12
(N
2
1)
Relation between Measures of Dispersion
Quartile Deviation (Q.D.) =
2
3
o or . 674S o
Mean Deviation (M.D.) =
4
5
o or . 7979 o
Coefficient of Variation
Coefficient of Variation =
c
X
1uu
Variance
Variance = o
2
=
_(X X
)
2
N
Variance = o
2
= _
_]d
2
N
[
_]d
N
2
_ i
2
P a g e  9
Mathematical Formulae
Skewness, Moments and Kurtosis
Karl Pearsons Coefficient of Skewness
Karl Pearsons Coefficient of Skewness (Sk
p
) =
Mcun (X
)Modc(Mo)
Stundud cuton (c)
=
Mcun  (3Mcdun 2 Mcun)
Stundud cuton
=
Mcun  3Mcdun+ 2 Mcun
Stundud cuton
=
3Mcun  3Mcdun
Stundud cuton
=
3(Mcun  Mcdun)
Stundud cuton
Bowleys Coefficient of Skewness
Bowleys Coefficient of Skewness (Sk
B
) =
3
+
1
 2 Mcdun
3

1
=
(
3
Mcdun)(Mcdun
1
)
(
3
Mcdun)+(Mcdun
1
)
Kellys Coefficient of Skewness
Kellys Coefficient of Skewness (Sk
K
) =
P
90
+ P
10
 2 Mcdun
P
90
 P
10
Kellys Coefficient of Skewness (Sk
K
) =
9
+
1
 2 Mcdun
9

1
Moments
Moments about Mean
1
=
_(XX
)
N
or
_x
N
2
=
_(XX
)
2
N
or
_x
2
N
P a g e  12
Mathematical Formulae
Correlation Analysis
Karl Pearsons Coefficient of Correlation or Product Moment Coefficient of Correlation:
When deviation are taken from Actual Mean
Coefficient of Correlation (r) =
_x
Nc
x
c
j
Where:
x = (X X
) & y = ( 
)
N = Numbcr o Poirs o 0bscr:otions
o
x
= StonJorJ c:iotion o Scrics X
o
= _
2
N
/ Covariance =
_x
N
Also
Coefficient of Correlation (r) =
_x
_x
2
_
2
Where:
x = (X X
) & y = ( 
)
Direct Method of Finding out Correlation
Coefficient of Correlation (r) =
_X
_X_Y
N
N
_
_X
2
N
[
_X
N
2
_
_Y
2
N
[
_Y
N
2
=
N_XY_X_Y
N
N
_
N_X
2
(_X)
2
N
2
_
N_Y
2
(_Y)
2
N
2
P a g e  11
Mathematical Formulae
Also the First Four Moments about Zero
F
1
= p
1
i
+A F
2
= p
2
+ (I
1
)
2
F
3
= p
3
+ SI
2
I
1
2(I
1
)
3
F
4
= p
4
+ 4I
3
I
1
6I
2
(I
1
)
2
+S(I
1
)
4
Skewness and Kurtosis
A measure of Skewness is obtained by making use of the third moment and second moment about the
mean. When the method of moment is applied, 1 is used as a relative measure of skewness. 1 is defined
by the formula:
1
=
3
2
2
3
or
1
=
2
3
, This is symbolized by y
1
. The value of
1
(beta one) shall be zero for
perfectly symmetrical distribution. The greater the value of
1
(beta one) the more skewed is the
distribution. If the value of
1
(beta one) is negative than the distribution is negatively skewed. If the
value of
1
(beta one) is positive than the distribution is positively skewed.
Measure of Kurtosis is the value of the coefficient of
2
. It is defined by the formula:
2
=
2
2
for a normal curve the value of
2
= 3 then the curve is Mesokurtic!. When the value
of
2
> S then the curve is Leptokurtic!. When the value of
2
< S then the curve is Platykurtic. On
the other hand y
2
is calculated that is:
y
2
=
2
 3
If the value of y
2
positive than the curve is Leptokurtic, If the value of y
2
negative than the curve is
Platykurtic and if it is equal to zero than it is said to be normal curve that is Mesokurtic.
P a g e  10
Mathematical Formulae
3
=
_(XX
)
3
N
or
_x
3
N
4
=
_(XX
)
4
N
or
_x
4
N
When Frequency distribution are given
1
=
_](XX
)
N
or
_]x
N
2
=
_](XX
)
2
N
or
_]x
2
N
3
=
_](XX
)
3
N
or
_]x
3
N
4
=
_](XX
)
4
N
or
_]x
4
N
Moments about Arbitary Origin
1
i
=
_(XA)
N
2
i
=
_(XA)
2
N
3
i
=
_(XA)
3
N
4
i
=
_(XA)
4
N
When Frequency Distribution are given
1
i
=
_](XA)
N
or
_]d
N
2
i
=
_](XA)
2
N
or
_]d
2
N
3
i
=
_](XA)
3
N
or
_]d
3
N
4
i
=
_](XA)
4
N
or
_]d
4
N
Converts of Moments about an Arbitary Origin into Moments about Mean or Central Moments
1
= p
1
i
p
1
i
2
= p
2
i
(p
1
i
)
2
3
= p
3
i
Sp
2
i
p
1
i
+ (p
1
i
)
3
4
= p
4
i
4p
3
i
p
1
i
+6p
2
i
(p
1
i
)
2
S(p
1
i
)
4
Moments about Zero
F
1
=
_]X
N
F
2
=
_]X
2
N
F
3
=
_]X
3
N
F
4
=
_]X
4
N