sur sus ccuims oF uiuepenaent variable.
Example 13.2: The following table gives indices of industrial production and
number of registered unemployed people (in lakh). Calculate the value of the
1992 1993 1994 1995 1996 1997 1998
107 105 «11210899
correlation coefficient.
1991
12 12 1986
Year :
Index of Production : 100 102 104
See ae 15 ered
Number Unemployed :
Solution: Calculations of Karl Pearson's correlation coefficient are shown in the
table below:18 175
119 18s
19.20 195
20-21 205
have
alues in the formula of Karl Peuson'scoreation coetha
Substituting values
3.2819 ip moderately negative,
eines correlation between ee
9 tnd nese Produit Henn mb
&
and aes OS ave tee BIE es the jcmttoton of items of production
SNS Sorrelaion ct snd eet og
‘Ste-group, ‘16 teay NAB 18-15 eg 20-21
No. of items 2g me Nag EO ata
Ne-ofdcteive tems: gg 72 "70 go gy SO
UDA Un com 1999)
yearn OUP ae be dented by ssi and number of deter
acetble > Caltlatons fr nan ai ‘comelaion
below:
a a)
foe Bm
Booed onl 2
ie we 3
add, ayes)
Pred -cay rg636.
* Fos Fon
6x -106— S14
fsxi9—(aF Jono a8)
949
ee valy
HeBatve,"Petncen A OF is negate, and is merely close € 1, Sati
item Ne Me bine i
% den, Conclude SPOUp) any (percent of defective items) is moderate and
& Tees 1 hat when size te eases, the number of delective
dae vera
the number of
¥ The fatto
Teported sei!NiNg data relate to age of employees an
En, sick
oye ick in 8 month
Wee yk eee are
Sick day, “7 82 35 40 gn 50,255
TOA
Bg a eee 8
Caleta Lo
te Karl
Kart Pearson's coefficient of correlation and interpret it
Clttion: [Kashuir Unie, BCom, 1997)
o
Calculation BE 4 sick days be rpresemed by varies «and. Fexpeciey
Mons fr vale of cnclaton int are show bw
- ‘Sick days:
Bate as 5.4 44
4 76 256 1 48
S98 56
a a pa 2 ae
4 28 36 5 <6
bs 2 4 2 4
50 4 16 4 2
be 6 36 6 in
35 9 81 5 2
a7 in 121 7 s
ao 6 sk oa
a ee
#
Ex. 0 2 sand y= = 4
aon 210
Substituting values in the formula of Karl Pearson's correlation coecent we
have
(2) Eh) 10230
Edy,
roxio /10xoF
Tracy? nb - OF
ti. = 0870
368
therefore age of employees and number
ce value Ot oa sh degree. Hence we conclude that 3s
terpretatin ine wed to a h
or posite cree eM gon ack eave more oflen cat
ng to the marks
requency aecordin
ine relationship
Measure the degree of
lowing table gives the
‘The following 17 Jence test.nea ee
% Sap,
——
mae eae
ae
ieee
5 ‘ 1 4
i:
7 3 ‘ 0
Pia ts
tt : ee
i
see ae a. 1 y5a5? =
1-2. 1 - 04857 = 0.5148
1 0
The rh stow mee
ot waiens a wo aan
posing cmeon eee peromane
Example 18.10: Jesks fel ths
vo fm as 12 clerks The longsering leks fee ak G5)
er ee Ne on ag free
Hier ter eit ett kL aaaereeer tor
rau, An me of in office. he shown ow eer
Ranking aecording
fem ee eae re
Ranking according
to effceny
gekstopea tebe ise alors e
rs li for seniority incement?
(Subbadin Unis, MBA, 1991
ranks ae aed ginen,caleaon for rank correlation coe
‘Do the data support the cl
Solution: Since
are shown below
Tak Acting Rast Acorn €
Tinad of Seow Bie)
& fe
T 2 7
2 3 s 1
5 5 oe ‘
4 1 3 4
5 . “4 6
6 0 m4 16
7 un a4 6
* 8 4 6
° 8 1 1
0 7 3 °
i 6 5 2exe
wet
7 aan” came
The ret shows alow depee pone corelation tires, eg of see
sd ftom she ain fhe tsk rs senate based om Teh
Service mo june,
Example 18.11: Ten competitors in a beauty contest are rahe by thre judges i
the following order
foie 1 Ogee a 4 7
pee er ere ee eer ee
jae OE 8k 8 TT
{Use the rank correlation coelicent to determine which pair of judge a the
‘nearest approach fo common tastes In bea. (MD Unis, MBA, 1996)
Solution: The pair of judy
‘beauty can be obtained im
(0) Judge 1 and judge 2
(3) Judge 2 and judge 3.
(68) Judge 8 and judge 1
Calculations for comparing their ranking are shown below:
eacoisiee Se icieve ea
Julee fale? alge Pa RF Pa OP Ao RP
rho have the nearest approach to common taste #8
93 ways a flows
Ry Re Ry
Toe | 8 7 7 w
gens 1 1 ‘
By = See ty 8 1 6
we 8 36 6 4
Stree 16 36 4
ee ae 6 4 °
ae 4 1, 1
O25 oA Ben 4 ar 1
ee 1 1 4
Hee ete et 1 pik 4
Eat = 200 fom to
Applying the formula
ox Sooo _ 1200
> Toao0-5 * '~ “o90 bet
6x24, 12
5 ~ OaBIe wy BH
Rant SGD Toavo= 09 70887
od 6x60 360
Faten 7!” oqo ~*~ ago ~ 88°
Since the correlation coefficient Ry ~ 0.636 is largest, the judges 1 and 3 have
nearest approach to common tastes in beauty.
(Cane 2: Whan Ranks are nat GivenSolution: Let
shown bel -
on 2 a8 5
Applying the formula R= 1-7
wi 22 air = 0107 0007,
336
jon between the debenture
The result shows a low degree of negative correl
prices and share prices of a certain joint stock company
1 if there was any relationship between
‘gathered from 7 countries
Example 13.13 An economist wanted to find ot!
unemployment rate in a country and its inflation rate, Data
the year 2004 are given below
Couniry Unemployment Tnfltion
Rate (Percent) Rate (Percent) _
x 70 3
B 85 82
c 55. 94
D 08 5.1
E 73 10.1
F 38 78
c 21 47
Find the degree of linear association between a country’s unemployment rate and its
level of inflationyes Since there are
sro the a each oF the te
ee Tone aie gh ae
Sotadon exw start ranking Ph
tran, thes
terns sb
: is
: = pence y= 2) and ales Sand
Tome sarcnncn speed Spine
(m3 =}
ona be noted hat. Her
ad oie, 9 2 ae
iodo tis
tsotvaatiey abo reper Psi
ak nt 48,
Rel
1 gio bee
iz’
79D
ofa
.gs75 = 0.0025
4514052 05¢05) = 1 -
dnvemory tornover and
“The result shows a very week postive anacaion betwee
coeficient between the variables x and
curingigs per sare
Example 1.15: Obusn the rank correlation
vals
Froese etlosng paso ere
eae vaste, a5 7700 “sot 8” 78 7
tale vis ras vo 5 130 120,115,160
’ [Mangalore Univ., BCom, 1997)
the variables. Moreover,
Solution: Let us start ranking from lowest value for bot
cele observations in both set of data are repeated, the ranking is done in
cordance with suitable average value as shown below.
Teale Rank ore ‘Rank —ifforne = RRP
sm fy d= R=
Wate faa AW cca RS oa baa, 15
4s 80 a
® 7 1s 4 3 as
ok eT 50 bored
me 48 Mo as 2035
ot a 5 pe
mo 2 108 60 Bred
75 wo 6 n
mo 9 sk Ee 225
0 0 25.00
o9 00.00
Je may be noted that for variable x, 50 is repeated thrice (m, = 3), 58 is
remedies = 2) and 65 is repented twice (my = 2). Also for variable, 110
peated twice (my = 2) and 115 thie (my = 3). Applying the forelsR=1
et
e BOD FO-a Le a+o°-s|
1000 =i
ata SUSIS2+05 +05 +0549) _ 1 — 8x1995 _ | 887
‘990 *
= 1 ~ 0.845 = 0.155
1- 3
00 990
The result shows a weak positive association between variables x and 9,
15.5.6 Method of Loast-Squares