Modeule tt!

geQupdo,
Meauras o Cerdal endeny MkKapoor
Afrnquray dist..bulen n genevat,
b the data QAur Aome Qersral value
DHeresl ratbodu gve disßevent qverqu ohich ae knech
as tho neasles o Conhal ferdene. îhe comonly e
eaeaes o cerlyal Val. Mean, Mecian, Mede )

Geoclie Meab and Harrnonie Maan.

Maan
e a det nvalues ed
Variable, theo the avithmeie ean CMean) gven

n
n be the
na
md. valun od the clan - interval having Grequenea
we have

Calaulaic o Mean
eupecialy ohen
Direel methd o cooputing caleulater
aplied to greuped dada involves Qellowing
thee, the
and in orcy to avoid
uied.
geheraly
0. 8her. oul Me thed
X: A+
Sfi
L sap devialion Mthed
X:A +h sf:u;
an

whee d=a-A and u: Q-4) Abig

elau indeyval.
and . the tqual
onbiag
(2) Maedean

H the valus o a vaviable are

the median

qunl to tf,
the
eveh Dh the madian
mid value

Hor the groupad data,

Median L+

ohare median clas

the
L:Aouer lioit o
N: total raqueney
frequahey o the median clas
f
h: widR the median clas
clan
c cumulatie Sraguny upto th
the median clas.
precsing
Quartiles: divida Hh
thale Valu hi
which
Guantile
Para, hn the valus!
into Gour equal part,
in Ehe ordar ob ragrila
arangd
 ovna.
od the uppar
data,

C),h

the clay
upio the clas
turulabive 4uguny or
6, , la
ohiek
precoding te clam in
ord lgwer quarku.
batuaon
baluaon te ugpar
he uppar
ihe dilona the

(9 Mode Vols o the

The rooda u donnd.
Vaviable
t the rotiou
data,
LLouey linit o the clas
claieontoiig
mode
czcew os modal reguthey
preading clas .

medal

clas,
o nodal: clas
h
4) Geometi Maan'
, g,) observalaen
the G-M. (14:.2n)

(agtlgst+ lggan) 0
logG M h

dittributoo, let ,, l2,,n be

n a

the Central valus, wit torauponding Aronuentis

wohure'
G.M

G-M=

(;) t
ti+thlog
 shou that
ence,
AM o
oy loq valuan.
loq os GM-
5) HarMonie Mean.

,22,,Xh be a set od nobservalaone

dasined asthe
the harmonie mean

Caritkmetic) mean oo the Teiprocay

Teciprocal o the
of the quanities Thus

HME
-
Sn a Srequerey dlistri bulion,
H-M=
, ohere
638 HIGHER ENGINEERING MATHEMATICS
Example 21-2. The following is the frequency distribution of a random sample of weekly
earnings of 509 employees :
Weekly earnings : 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
No. of employees : 3 6 10 15 24 42 75 90 79 55 36 26 19 13 9 7

Calculate the average weekly earnings. (Rewa, 1994)

Weekly Mid value No. of Step deviations
earnings employees
f fx* u=(-25)/2 fxu
10--12 11 -7
3 33 -21
12-14 13 6 78 -6 -36
14-16 15 10 150 -5 -50
16-18 17 15 255 -4 -60
18-20 19 24 456 -3 -72
20-22 21 42 882 -2 -84
22--24 23 75 1725 -1 -75
24-26e 25 90 2250 -398
26-28 27 79 2133 1 79
2830 29 55 1595 2 110
30--32 31 36 1116 3 108
32-34 33 26 858 4 104
34-36 35 19 665 5 95
36-38 37 13 481 6 78
3840 39 351 7 63
40- 42 41 7 287 8 56
+693
Ef- 509 Eft = 13,315
Efu= 295
 by dinect mthod,
13316
X

By sap "}evialcs' nethod,

 Calculat median and the loue ard
asple:
digtributon o mavke
uppe quar tileu er the
Sei
obAaibed by'h9 Budent. ind alo the
Ob£aned
and the 'mode
inte quas ile Tag
clas Qulabve requehey
(Marta qrou) CNo. o studra) (Las
than) CMore Hren)
5
6
I5

36
I5-3 "13
30 - 36
36- 0
Joln: Hedianor&,

Macdion LN-¬)

Madian (4 ) Salts in the cas (15-20) ard

given by
"is+(-) x5 t6+ 13-6 3

= I9-6 orks

oueY Quar Hle,

fatte in the ela

4 e ( , 2a5) alo atls

CI5-20) and by.

.Qe5+ ( . n) K: 15+
5 lQ"5
3

Quon le, : L (N-c).

36.15) ats In Ehe clan (35-30)

by
36)
Q = Qs +(3615-
5

Bei- Qaantle Rarg @-)

+- &.)
Mode =
At

the claus. tr5-0).

Mbde Value ala in

5-6 X 5i
l5 +
,. Mode = (15-4)+(5-10)
and 25-7 (for froquency 3876).
a certai
Kxample 21-4, Given below are the marks obtained by a bateh of 20 students in
class test in statintics andl economics
Roll No. of Marks in Murks in Roll No. of Murhs in
Maths
Marks in
student8 Maths Bconomics student Economics
63 58 11 26 10
54 66 12 42 42
52 26 13 33 16
4 32 32 14 48 46
30 26 16 72 60
6 60 85 16 61 64
7 47 44 17 46 39
8 46 80 18 33 38
9 36 33 19 66 30
10 28 72 20 29 36

In which subject is the level of knowledge of the students higher ?

The subject for which the value of the median is higher will be the subject in which the level
of knowledge of the students is higher. To find the median in each case, we arrange the marks
in ascending order of magnitude:
Marks in Marks in Marks in Marks in
Sr. No. Sr. No.
Statistics Economics Statistics Economics
1 25 10 11 46 42
2 28 15 12 47 44
3 29 26 13 48 46
4 30 26 14 51 50
5 32 30 15 52 55
33 32 16 53 58
7 33 33 17 54 64
8 36 36 18 60 72
9 42 38 19 65 80
10 45 39 20 72 85
STATISTICAL METHODS
641
Median marks in statistics A.M. of marks of 10th and 11th terms
45 + 46
= 45.5.
2
Median marks in Economics = AM. of narks of 10th and 11th
items.
39 + 42
= 40.5
2
Since the median marks in Statistics are reater than the median marks in
level of knowledge in Statistics is higher. BconomiCS; the
Examnle 18
Exampk Sruguany ditibulion
An incoplkt
bule:
60-o
So-60
80-40 Ho-50
Variable:
66
36 R
L
hequany
181
and
the total requeny
Given that
Qind ther mising regunii
Median

leln:
the
be the micing reguancia
ahd 50- 66
clases
the meian liis in the clas yo-5o,
3

29 - (9+ 30 +f1)
G6
:. 33-5
e34
Q99-( ta 430+ 34+65 +a5+18)
Erampe
The ctis A.B, c ae
othan A rofàt lanl fao> Ato B at 30 kmlk
frors B to c at o knlkr, roo Cto A at so kml k.
Beteaminä he avwAnge spsd.
AB BC=Che t kr

fine taken to liaelron Ato B sl30

Line laken "to aavel rorm Bto c slho
takan to tanel frorn c to
 km/fo. 38-8 =
) time iotal spud Bvg
Total
avuay4pddisfari th, Hene
3
speed arasgu
the Thus
 Meaunes o Dispovgion
he insporarl da person
toeasues are

Thà i the inplut meaure e duperson

Rarqs giver by tbe dierere betwer te
ahd the ditributeo
Ond the least values si
grataut
3S 39
93
Value.- Min. Value
Kange Man-

) Qluartle Bevialien Y inleraquark Rarze

and 3 ae tf i t and t d
quailes, the leroi- invoquast varge
Q)
3) Maan Doviation

tta tooole.

4) lardard Devialcn (3.2)

D)
whne N u
 Note ro any
Sror
T4 the dlevialiens are n asured
ibslead of H u called
bay A.
Othey value, A,
deviatioh
Rquare
The voot - toean
slandavd devialon
The Squaveof the
Vasiabe
Kroon as the

Calculatien o sD.

£ 8hort- cut oe thedd

Sfi sfiJ

D s tep- deviaten athud

Lfidi

wbere di= x;-A ard d's i

toeas Squase devialon u

The root
by Sfidi
(a-A
 Coedicient od Variation
The Yatio ot the starolard doviaain
to the the variato
Vaiaton lomparing
u uwod Jordilerest
o-6 tk
befoeen two 9roupe oi \0
Coodticeht ot Vavialion =
Rolations batween Neaures

li) Quarile deviaton - & Cttardarcd

3
cde via ton)
Ci) Mean deviation hstardavd deialen)
5
cornbinaon od
) tahdavd Aeviaten ot the

the me ah
where D; = mi-m,m bebg
the bibed
cosbibed arsple
 mean and S for
Nahople O, Palealate th
9
I3

dolo:
fxd

-3 -9
6
-9

3 36

Sf- 48 Sfl: o

MeanA+ 2fdi

fi Sfi
 foltowin
Aaleulal the ma.ab abd 8A ot the
fequoncy dirfibuticn:
Weky Wagu in ood os ib R
R
3
725- 05 Sas-6o"s

6o5-&8s
d&5- 36s I8
6&s-6"5
36-5-Hhs 5

A: 3a5, h: &
Wagu clau |Hid
Rs fol'
-3 36
la:5-05
do-s- 8-5 -

A85-3b-5(325)
265-A4S

SA5-6o5 5 45
6o5-68-s1
3
50

fa' -13 393

 Mean X A+ hsfa'

39-5 + 8
()

|393 -13)

= Rs l6. 64
ate
Beores et
Exatopl the fotevng
o innig!
a Seaies
balie A ard B io
36 &H
13 19 19
15
A
87 48 13
51
l6
B
getfes and obo u
the
botfer tore

Let debot &lo

and y dorot glore
 Take A- 51 as
dy-A ds
=2-61
-4
-39 -39
- 35
-H5
H
13
1936
C
-3d 5!
31 96
68 4624
86 -I5 - 38
33 13

LT508 9309

Tos A,

I150t lo

-( )
Varia uop htg x to0 = &3-6'%

y: A+ S de 51 + -240
h
 Sdi -C)

69.6%
A A tB,It oL loos that
ipce tbe AM g
A u better seore getter.
Maiaton tt B lhu
Sibe tbe do beab

IBat B iu
X Example 21-10. The numbers examined, the mean weight and S.D. in each group of
examination by three medical examiners are given below. Find the mean weight and S.D. of the
entire data when grouped together.
Med. Exam. No. Examined Mean Wt. (bs.) S.D. (bs.)
A 50 113 6
B 60 120 7
C 90 115 8
(Mvsore. 1994)
kolo:
Ge, D=50, y = 13, o,6
Pa6o,

(itbat ng)
DDi+ D¡ D, + b3 D3/
ohere
hWe bave to dend abd

Da l15-t16--)
C?0X64
O> 6o t bo4 qo) kox6)+ (6ox 1)+ C7°4

I900 o
1

ence V6o
 Momol
Morvena about Wean ( Corssal onent)
th
are the Value the hr
deqirad au

,M3,, n ae tk Yalueu woith requencea

)
n

N
N
to parteular,
Vasiabe
N
 (A) ()
Momsrt about an
S (;- 4)
N

S { (i-A)
N

the Caleulaton by
Fase
h whare
N
irdividua geriu
No er an
S_(i-A)'

=
 227
Moroar about the n
n
whe
zz
V
n

N i

N i
ard so on

N
Rolaiin behoar and

RJaton beluesn Vy and

Relalin

(
M+ 3
 karl
Pearsons B and Coalicianl

p-oeticien

tiree nnorsenk dictnion,

Variabk
variabe are l,l6, -4o.
od the
the
about the Value Variane is 15 ard
ean is 3,
&hoo tat te
Rheo
Hg -86. =-H0.
Gn, A=2, Ui=

3
+

=-26.
 a
Example 5. The first four moments of distribution, about the value '35
35 are -
1020and
144000. Find the values
v of Hg H3» Ha 1.8,240
"H, = 0..
Sol. 240 -(-1.8) = 236.76
H, =
u, = L,- 3u,u+ 2uf =- 1020 - 3(240)-1.8) + 2(-18)39e
3

=H 4uu + 6uu-3u4
= 144000 4(- 1020)( 1.8) + 6(240)\ 1.8)² 3(- 1.8)4 = 141200 3.
lculate the variance and third central moment from the follou
 Q3 calulak he
the Variaba ard the third
Cantval monent frors the fotlourg data
6
1

solo:

fu3
fu
h

-1

-69
3 69

51

58 ||6
6 a9 189
1
I6
J4
5o -37
N=Sf:
A56 fusotS
 hira CeII

Beamnle 7. The first three moments OT a distribution about the value 2 of the orink).
m1 16Rnd-40respectively. Find the values of thefirst three moments about the origin.
Sol. We have A= 2, Hí=1, =16, H,=-40
V =t A#Lf= 2+l=3
V, =ly t F= 15 + (3) = 24
V, =H, + 3u,F + 7=-86 + 3(15)(3) + (3)3= 76.
Example 8. The first four moments of adistribution about x = 2 are 1, 2.5, 5.5 and 16.
Caleulate the first four moments about the mean and about origin.
H= 5.5, Hi= 16.
Sol. We have A=2, =1, H==2.5,
Moments about mean
H=0
Hy=ý- (4)² =2.5 - (1 = 1.5
*y-3uyu +2(u,)9 =5.5 -3(2.5X1) +2(1)³ = 0
Hy=í- 4u4 +6u4{ = 16 - 4(5.5X1) + 6(2.5X12-3(1)=6.
Moments about origin
V =* A+ Wj
V3 =H3 +3u,7 + 7', =Myt 4u,7 + 6u4,F?+ F*
V=* = 2+1= 3, V 10.5
V, = 1.5+ (3) = 6(1.5X3)2+(3)=168.
V3
=0+3(1.5X3) + (3) = 40.5, V4
=6+4(0X3) +
 Q6er a a diebibukop, H. oean is (0, Varianea
Aod the
Sour oroanl about the eriqib
Geiven, a= 0,
1

,=
Pa
Geo,
Wki, =}
3
3
(16)
 dluo, Given Ba=

Mormenk abaut onigin.

V= =l0

+°
V8 M3 + 3M
= 154h
 VExample10, In a certain distribution, the firstfour moments about the point..
-L.5, 17,-30 and 308. Find the moments about mean and about origin..Also,
Also, caleulate B,and
Sol. We have, A=4, , =- 1.5,, = 17, #'; =- 30, #', =308
Moments about mean
(U.PTU.2014
Hy ='-H= 17-(- 1.5) = 14.75
Ha= -3u',#,+2u=-30 -3 (17) (- 1.5) + 2(- 1.5) -39 73
H=P4- 4u', 6
+ -3u
=308 4(-30) (- 1.5) +6(17) (- 1.5) - 3(- l.5) =342.312%
Moments about origin
V,= =+A =- 1.5 + 4 = 2.5
V, 4, +t 2 = 14.75 + (2.5² =21

V; = + 3u,F + = 166

V4 = Ht 44,7 +6u, + = 1332

Calculation of B, and B,
B, = = 0.492377 B,= = 1.573398
Example 11, The first four moments ofa distribution about the value '4' of the varat
are - 1.5, 17,- 30 and 108. Find the moments about mean, about origin ; B, and B, Als, fud
(A.K.T.U.2017
the moments about the pointx = 2.
STATISTICAL TECHNIQUES - 215

sal. We have A 4, 1 - li,2 = 17, u: =-30, u',4 = 108

Moments about mean

Hg= - =14.75
Hg = ' - 3u', p, + 2u = 39.75
Hy =l4-4u, l1+ 6u' u -3u = 142.3125
Also, F = u1+A=-1.5 + 4= 2.5
Moments about origin
V,= * = 2.5
V, =uo + x = 14.75 + (2.5) = 21
Vg = + 3u, + * = 166
V4 = H4 + 4ug + 6u7 + = 1132
Calculation of B, and Ba
2

B1 =
3
= 0.492377 B, = = 0.654122
Moments about the point x = 2
=x-A = 2.5 -2 = 0.5
= 15
l = lg t H, = 14.75 + (.5)
W= h + 3u', u,-2u,=39.75 + 3 (15) (.5) -2(.5) = 62
= 244
4= H + 4u',u-6u'uf'+ 3u
thedistribution is negatively (posit

0 Karl pearsons Neth od

)
3)
Bousleuk methad
yettod
Kelly'
 ) Kavd Pearsos nethd
Karl pearsens cbesliciert et gkawneM A M- Mode
S.D

Mode 3 Median- AM
karl Peavsons coecticiant ot &kuonas
3 Median a AM)
A:M-

3 A:M -3 Nedian

- 3(B- NMedian)

Kavl peason't lceficient af

derstd by Skp Nean- Md
Skp =0
Mean Mook
Distsìbuion is sy mneaical

Mean - Mooeso
( Mean
Mean s Moda
Disibufen is poalively skuwrd.
Mear - Mooe Meah < Moda
t3) 2f Skp <o
Dsibuton
hugatnaly s2aued.
 3
98816 Medias=
e0 Median- 98-816
3
100-18H: Oole H
ode lo0-M 0
iahe 8.D=
ode - Skp Now,
0 8kp= 35, Variabte= t00, = Gu solo
abd mod o-.icod
ia diab me
upes keuo Cofferest
af
poarso
br karl abd I8
35 Vayiahce the to0,
te the data, skewd oclesalely 2Joa
65
0°32= Mode - Skp= buo,
ole M96 Gun,
a9.6 = 65, 8D-= 034, SKp=
the omoodo the find 6.a9- 4
65 deviato &tabclard 034, dislibution
is
SkeoneH
a o Coeficient
iebt Poaona Kar l Q1
 arlaio diehikileo, th Qottouoe Yeatk
Were obt£ned:
data, ailed to ae
person wko Jave. ou thu
the value of SD. You are Tequined to esti at
twit tfe halp of availabte dada

Aset Skp = -0 , =46, Modian hg.

Cth,
Noo, Skp = g ( - Median)

-04= 2 (46-48 9

ohervalaohs is 300 abd &u

is Soo0 he oecdian u I5. ticol
T6e
tBe Karl pasos oeffeiort of Skure
3%=3o0, Sx s0o0, oediabt
b=l0,
360 I5
ow,
So00_t6 5
h Jo
3 (-median)
(a- ediah) 2 (15 - I5)
Now, Skp =
S. D
 S.D. Þ.94 Ö.J4
Pearson's
t's coefficient of skewness for
Example 6. Calculate Karl the
Income (in ) 500-600 600-700 700-800 800-900 fo
900-1000
l owing data
No. ofemployees 8 12 4 2 1
10 0-10 1
Calculation of , Mode, S.D.
Sol.
No. of Midpoints d=x-A u= dlh fu
Income
of classes A = 750 h= 100
(in ) employees

8 550 - 200 -2 -16

500-600 32
600-700 12 650 - 100 -1 - 12 12
700-800 4 750 0 0

800-900 2 850 100 1 2 2

900--1000 1 950 200 2 2 4

1000-1100 1 1050 300 3 9

N=28 fu =-21 Efu'z 5958

g=
 -I
TICAL TECHNIQUES
STATIS 229

Efu 21
A.M. N |h=750+
+ 28
(100) = 750 - 75 =675
Mode. By inspection, modal class is 600-700

Mode =l+

Here l= 600, A,= 12- 8 = 4, A, =12 -4=8, h = 100

4
Mode = 600 + (100) = 600 + 33.33 = 633.33
4 +8

2 2
fu 59 21
S.D. = N
-

( xh=
28 28
x 100

= 2.1071- 0.5625 x 100 = 15446 x 100 = 1.2428 x 100 =7124.28

X- Mode 675 -633.33
Now, Karl Pearson'scoeff. of skewness = S.D.
= 0.34.
124.28
Method o Moroen /0

Cab be
Morsent lotient of gkeObes ear
Writes

The morerst Cofiient ot gkaoy u albo

denod ,
The Rxt threa Cebtral momebk of a
digtibution ae 0, I6,- 31. ind the
Coefficiernt ef gkeuoeu
Hi0, MaI6, M3 -31.
btn,
toeticierst f &keuhess =
Mornent
 fous
Qre , Q0,
about the Value 5 of the Variable eootfieiaot
Caleulaa the roroent of
gkwwness
solo' =50
Gtn,
Mornent coetficiert gkwne
NH

Merert toe of gkewres

ke NU)3
 of skewness for the following
Example 3. Calculate the moment coefficient
Classes 2.5-7.5

8
7.5-12.5

15
12.5-17.5

20
17.5-22.5

32
22.5-27.5
23
27.5-32.5
17
dis32.trib5-ut37.io5n:
Frequency 5
Sol. Calculation of Moment Coefficient of Skewness
Classes Mid-pts. d=x-A u = d/h fu
A = 20
fu2
h=5 fu'
2.5-7.5 - 15 -3 -24
72
7.5-12.5 15 10 - 10 -2 -30 216
60
12.5-17.5 20 15 -5 -1 -120
20 20
17.5-22.5 32 20 20

22.5-27.5 23 25 5 1 23 23
27.5-32.5 17 30 10 2
23
34 68
32.5-37.5 5 35 15 136
05 45
N= 120 135
Efu =-2 Efu' = 288 fu -62
-2
Now, 5 =-0.083
120
Efu2 288
N 5 = 60
120

-62
N 5 =-64.583
162
Now, Hy = Hy- u= 60-(-0.083) =59.993
Hg = H -3u + 2u=64.583-3-0.083)(60) +2(-0.083
=-49,644.

Moment coefficient of skewness = 49.644

=-0.1068.
Vu J(59.993)
 Kurtosis .

distribuion te leptokuvkic
distributon is mesokur
) B¡ >3, the is pbytykuvte
Ba3, the
distnbution abo moaured
3) pa <3, the aloo
Ba hibution is
a die
The kurtosis op , B-3 .
as
ohich iu dadired leptokurfi
by >3
oeokurte
Pa<3, Platgkutie
Ya 20A-840
abeut ano
four onent 35000-Diseus
ho Qrst
Q.) he dictibuo ge o, (00,-1 and
distibertion.
Srequargy,
kurfasis be
the ua=and Hy= 35oo0
Guven, Mi=0, Us lo0,
Now, Pa: Hy 35000

The olislbibuton i lepto kus tie.

 Dyample 2. The first four moments of a distribution, about the value 4' of the variable
are 30 and 108. State whether the distribution is leptokurtic or platykurtic.
are-1.5,17,-
(U.P.T.U. 2014)
Sol. We have, =-1.5, u, =17,u=-30, =108
Moments about mean:
H = y-H=17-(-1.5)² = 14.75

- 108 4(-30)(-1.5) + 6 (17)(-1.5)2-3(-15)4= 142.3125

142.3125
Kurtosis: B, = H42 (14.75)2
=0.6541

Since B, < 3, the distribution is platykurtic.

 distibu
blain the various
Qboeut I, , lo abd h5.
havaclordies f sibution ob the
Gives ifosrmalon Cooment upor
baturebd the distributioh
Aolb Gtb,

Morer loag4iuent

3
3

4
-3

tis u dyronoetieal.
The distibu
Kutosú
palkurte
QH the stabdard oda vialon of a sym mele
oistribulon à 5 . Nbat rust be the value o the
momest abot the hea èo 0rles tfat the
distibuton be
ti) lepto kus tie Cn) mosokuhe Ci) plalkus he ?
Solo
we bare 5 > o=5
25 > 5
Now,
H
C) leptokurtie
6A5
(iD Meso kustic B 3
L 1815/
ci) plalgker he Bas
23
 (ii) Platykurtic if P2 625
moments about the working mean 28.5 of a
Also dievaluate
stribution8.
Example 5. The first four
and 454.98. Calculate the moments about the mean. Also
0.294. 7.144. 42.409 kurtosis of the distribution.
and comment upon the
Sol. We have,
Moments about mean
skewness and
H= 0.294, L, = 7.144, ,= 42.409, u=454.98 (A.KTU.2016
Hy = -u=7.144 - (.294) =7.0576

= 42.409 - 3(7.144) (.294) + 2 (.294) = 36.1588

= 454.98 - 4(42.409) (.294) + 6(7.144) (.294) -3(.294)4

= 408.7896
Calculation of B, and B,
B, = =3.7193 4 8.2070
=
3

Skewness
Since B, is positive, Y, =1.9285 is positive
:. The distribution is positively skewed.
Kurtosis
Since B, = 8.2070 > 3
The distribution is leptokurtic.
Example 6. The first four moments of adistribution about the value 0' are -0.20, 1.76,
-2.36 and 10.88. Find the moments about the mean and mneasure the
kurtosis.
Sol. We have, =-0.20, u', = 1.76, u', =-2.36, u', = 10.88
Moments about the mean:
B=0
H,=-4= 1.76 (-0.20 =1.72
-2.36 - 3(1.76) ( 0.20) + 2 (-0.20) =-1.32
 STA)TISTICALTECHNIQUES -| 237

= 10.88-4(-2.36) ( 0.20) + 6 (1.76) (-0.20)-3 (- 0.20)

=9.4096
Kurtosis:

= 3.180638

Since, B,>3 hence the distribution is leptokurtic.

Eeample 7. The following table represents the height ofa batch of 100 (A.K.T.U.
students. Caleu
2018)
latekurtosis.

Height(in cm) 59 61 63 65 67 69 71 73 75

No. of students 2 6 20 40 20 2 2

Sol. To calculate B,, we will have to first find the values of u, and u:
Moments about 67

1004= 0,24

Efu2 164
(4) =6.56
N
=100)
x-67
Height No. of fu fu? fu' fu
2
(cm) x students f
59 -4 0 0
61 2 -6 18 -54 162
63 6 -2 - 12 24 - 48 96
65 20 - 1 - 20 20 - 20 20
67 40 0 0 0
69 20 1 20 20 20 20
71 2 16 32 64 128
73 6 18 54 162
75 4 32 128 512

N= f= 100 Efu = 12 Zfu? = 164 Efu =144 fu = 1100

144
x8 = 11.52
N 100

1100
x 16 = 176
100
 ATEXTBOOK OF
238 ENGINEERING
Moments about mean
Mg = - =6.56 -
(.24)² = 6.5024
MATHEMATIC
- 4ujHí
H == 176-4 + 6 u - 3u4
: (0.24)= 167.19798
(11.52) (0.24) + 6(6.56) (0.24)2 -3

Kurtosis
= 3.9544 > 3
Measure of kurtosis B,
leptokurtic.
Hence, the distribution is
 Curve- ithig
Squave
line y= a+ ba )
tirg a Staigkt
2y= ba + bs
b
b
 55b I5a+ T48= 3>
bI55a+ -oH
68
55 5
9
b5 tere,
aSay
Sy= are ogrs Norral
ofinu stsaight the at
be best
55
: fotouoing the foa but that lirestaight
the fird quases, teast otthoad oe the By
 tBese
a-0, 6- /3. 6
 Hence
2. Fit a straight line tothe
Example 2.. following data by least square method:
1 2 3 4
1.8 3.3 4.5 6.3.
1
(U.K.T.U. 2011)
,Lot the straight line obtained from the given data be y= a + bx then the normal
eguationsare
}y = ma + b Ex ...(1)
}ry = aEx + bEx2 ...(2)
m =5
Here,

0 1 0
1 1.8 1.8 1
2 3.3 6.6 4
3 4.5 13.5 9
4 6.3 25.2 16

Ex = 10 }y = 16.9 Exy = 47.1 Ex =30

From (1) and (2), 16.9 = 5a + 10b

and 47.1 = 10a + 30b
Solving, we get a = 0.72, b = 1.33

Required line is y = 0.72 + 1.33 x.

(, Y,
 b
expohenfial Curve y- ae

y= A+ BX
y-lg.4
xy AsX+Bs x
2
Al9o
B-blog,e
e tbe curve y= axb
htfing
y= A+ BX Y=lg9
Sy-nA+ BS X A log oQ
BSx B=b
Sxy: Ax +
X= log
Curve 4 ab
the CuTve

y= A+B X y log 9
A= log a
2Y b A+ BX
B: log b
SXy = ASX + BSx
X
 o te eumre 9
4tirgVlog log k-
pv'k
P D+BSX
logSy=
,X
Y:log v, A: log k, B:

-31 be
ra o the cunte. Qe

-32
6 be
a Se
+bs e
Sye
Sirg th urve

}xy =a

the Cusve Y=

3yva- Co S
-30
af the Curve

3 6 AsSx
Sye a Se
+6Se
Sye= a Se
 d the curve but it of the tgpe
data:
= aebz to the olto utng

I6

u g- ac (or)
ted 8=Y= 0+Bx0
The earve to be
The rovol equations
BSx
2Y- DA +
X=
log, a , 8-b log,,
Gake Y= Log,9 . A-

k|1609|26 5 5-R805
6 -5544

(o.5849

L5-866
l"3222)

sY5.1536 SX* 300 Sxy: 4o-gg62

8862

SX: 34
 curve Taguirod the Hente
49 0-43
B
b
9-4154
A
bgT66,
B= A=
0-0956)
ghs, two thue
4o&868
300B At 34 =
34B 5A+ 5.1636
Subatuing
value he
and(
 Hence tl
Example 2. Determine the constants a and bby the method of least squares such that
fitsthe following data:
2 4 6 8 10
4.077 11.084 30.128 81.897 222.62

Sol. y= aebx

Taking log on both sides

log y = log a + bx log e
Y=A+ BX,
where, Y = logy, A= log a, B=b log,, e, X=x.
Normal equations are
EY = mA + BEX ...(1)

and EXY = AZX + BEX?. ..(2)

Here, m=5.
Table is as follows:
X Y XY X?

2 4.077 2 .61034 1,22068 4

4 11.084 4 1.04469 4.17876 16

6 30.128 1.47897 8.87382 36

8 81.897 8 1.91326 15.30608 64
10 222.62 10 2.347564 23.47564 100

EX= 30 EY =7.394824 XY =53.05498 SX' = 220

Substituting these values in equations (1) and (2), we get

7.394824 = 5A + 30B
and 53.05498 = 30A + 220B.
Solving, we get A = 0.1760594 and B = 0.2171509

a = antilog (A) = antilog (0.1760594) = 1.49989

and B 0.2171509
b= = 0,50001
log10 e 0.4342945
Hence the required equation is
y= l.49989 p0.5000 L*
 foT ab or lh
3 obtain a Telaion of the ors Square.
ma thod of leat
jotloujt data by the
5

331 65
83 I5-+

be t d à
itted ab
Curve
The
A+ Bx ohere
a ) 6
B loq b
The bora! eguaticons as

Bsx2()
AX +
Sxy

8 38a,
83 0.419
3-5616

3 I6-+ 6-0792
33-!
9-0tIo
A5
La-6312
652 36
6 sx'90sx 33 1812
Sy15 55

in ad
valu
Bh: thee
= 5A + 0 B
T6465 90B
loA +
33-t81a
tuo
thoe B:0:3
and
0-31
A = Qo4
anfilog A =
a:
B== |995
antlog B
b=
Teguired uYVe
Heree the
 Q. Obtain the leat iquaru t of tho orm
fct)= -3t
a+hot oy the data:

.16 O58 -3t

t
u fct): ae
Boln to be
the curve
The roma) euadicns aTe

-3t 6t+bse-5E
S.f(t) e =aSe
2L Seb.bs HA
a

-5E

O6922
bt03 0-6065
6703 5488
6 o3888
O-4493 6-3619 3012
3012 223) L653

O20{9 (3630-0901
35
L6221

abd .
Bubiall Values in ft)
4o98 - to6 a+ I: 3328 A

thue
sobing o.678, b=0-312|
eguireo cure
Hepce, the 3
-at I"/65
fCt)- o» 6118e -31ate
Q-5 By tha oet6ed o loast squates, ibd h
euve y:an+ bu? that but fia the foltou ing decla.
y|8|St89 98

sot The Cuave to be ditted u ysar+b.

The norhoal egha are -

5-| L6

89 81

64 A56 S6-H
99
824-9
56
q19

Thue tuwo,
Hence, the vegued
tt the Curve pV'=k to the totouwing data:
PCagle) 5 3
I000 150620|
6 540 60
Bolh: pv
) P
 qaking tog ,
doy )P
tog k' log
log P

whee Ylay , Xoq p. A log k, B- -

be rorma) equaliors are
Sy- D + BS X

P X-ley P
3-a095s Lo606 009062
0"gol08
3
50617lo03/0)
Q81506
T50 O840S o 09062
640 | 080got5936
d5 39 194 tl6o0
520 lato4b-A4
460 O.A 71 l2 Q665
3 659825
7a557313196

L7 SSST3= 6 B t os1S B
3/96 = loSUSA + O5994S B
tho tw0 g,
A-999/1, B= -0702 8
 B
wkT, AA log k
abtilog Ch 26629)

Hebe, the ruired

required cuve U
 Example 7. The pressure of the gas corresponding t0 various volumes Vis
given by the following data:
V
(cm'): 50 60 70 90 100
P (kg cm): 64.7 51.3 40.5 25.9 78
Fitthe data to the equation PVY= C. (A.KTU:
Sol. PVY= C
P= CV-Y

Taking log on both sides, we get

log P = log C -y log V
Y =A +BX
where, Y=log P, A=log C, B=-, X= log V
Normal equations are
2Y = mA + BXX
and EXY = AZX + BEX?
Here m=5
STATISTICALTECHNIQUES -| 255

below:
The table is as
P X= log V Y = log P XY X?

64.7 1.69897 1.81090 3.07666 2.88650

50

51.3 1.77815 1.71012 3.04085 3.16182

60
70 40.5 1.84510 1.60746 2.96592 3.40439

90 25.9 1.95424 1.41330 2.76193 3.81905

100 78.0 2 1.89209 3.78418 4

EX =9.27646 £Y =8.43387 XY = 15.62954 EX?= 17.27176

From Normal equations,we have

8.43387 = 5A + 9.27646 B
and 15.62954 = 9.27646 A+ 17.27176B
Solving these, we get
A=2.22476, B=-0.28997
Y=-B= 0.28997
CE antilog(A) =antilog (2.22476) = 167.78765
Hence, the required equation of curve is
Pyo.28997= 167.78765.

Example 8. (i)Given the following experimental values:

1 2 3
2 4 10 15

Fit by the method of Least squares a parabola of the type y =a +ba".

(ii) Find the Least squares fit of the form y =a, + a to the following data:
-1 0 1 2
2 5 3
fo ith noint ( ) is E. = (v. -hr 2)
256 ATEXTB00K OF
ENGINEERING
MATHEKT,
0 2
1 4 1 4 1
2 10 4 40 16
3 15 135 81
Total Erky = 179
Ey = 31 Ex? = 14 Eh=98
Here m =4
From (1) and (2), 31 = 4a + 146 and 179 = 14a + 98b
Solving for a and b, we get a=2.71, b= 1.44
Hence the required curve is y = 2.71 + 1.44 x".
(iü) Normal equations are
y =mao + a Zr ..1)
and
Er'y =a, Er + a, Ex ...2)
The table is as follows:

ty
-1 2 1 2 1
5 0 0
1 3 1 3 1
2 4 16

Ey = 10 Ex'=6 Ex'y = 5 Lt= 18

Here, m =4

From (1) and (2), 10 = 4a, + 6a,

5=6a, + 18a,
a, = 4.1667, a, =- 1.1111
Hence, the required curve is y = 4.1667 - 1.1111 y2
 to i t
Use the hood of least squa
tabk
the Carve y= Co + CVz to the ottouweng
of values :
6

to be Hed Co
The ceeve
are
The horral ege
Co

+ C S2
Co
 Table 1S as beI

1
ylx

21 210 6.64078 3.16228 100

0.1
11 55 4.91935 2.23607 25
0.2
0.4 7 17.5 4.42719 1.58114 6.25
6 12 4.24264 1.41421 4
0.5
1 5 5 5 1 1

2 6 3 8.48528 0.70711 0.25

1
Ex = 4.2 Ey/x) = 302.5 Eyyx = 33.71524 = 10.10081 = 136.5

From equations (1) and (2), we have

 thue toD gh,
Co = l9732 7,
Ci= 3 2818 Q.

The raguird g
32818 AV
y= 97327 +
 Corelaion
CoTrelatior.
Karl Peartons toedficient of

ngaoy
Alternat forr ot v(x9)

(S)' Sg-(sy)
Do. o pais ot value o
Here, b u the

00worelaiso eo-esficit u indeperdent

Aot and ea le .

change
onigin
Let
h
 rlu,v) DSuv usv
nsu'(suJosv(s)
corelalon for the
83 Aind the coeieient
conieiernt
^ottowtbg table:
30

6
86
(8

tHore b6.

LO
l68
L& 576 432
36 132

20
1A96 log0
36
3276

126)
6 CA772) - (IaDX
(120)N6x3a76- (U26)2
y6xabeo -
 fotloewerg percontage
got the Od olatisthea
pibeipl of Ecoronmiès
oavka in prinepla 415 8A9062|6529
Lasz
Roll boS? 3 39
Eeobormica 78 26 86 58/ 53 d
l60 l6e l646
8latisia &45t corre laton.heroly
coef od Abe eolod
Calceulal the
the taoo Bubjca be
Soln: Lot t6e ark èo
arol veapectinly 66
h

e uy
=X- 65
69 324d34
84 da5A85
-29 - I5
36 6a5825
98 36h0
-6 I660
68
82 289
90 hoo So0
6J5
6 68
-3 - 8
65 S3 - 13 l69
39 -l9 676 g6/
5398 d134,
0Suv- SusV
Yuy =
ur
yosu? (su)> Nosv (sv)'
= 0"789
19x 5398 -Co) Nttox as4)-Co) /
 Example 1. Find the coefficient of correlation between the values of xand y:
3 5 7 8 10

8 12 -15 17 18 20
y

Sol. Here, n = 6. The table is as follows:

y2

1 8 1 64
3 12 9 144 36
5 15 25 225 75
7 17 49 289 119
8 18 64 324 144

10 20 100 400 200

x= 34 Zy= 90 £x=248 Ey'= 1446 Exy = 582

Karl Pearson's coefficient of correlation is given by

n Exy - 2x Ey
r (x, y) =
Jnx? -(Ex² n y? -(Ey)?
(6 x 582) - (34 × 90)
= 0.9879
J(6x 248) -(34) 6× 1446) (90)
Example 2. The following data regarding the heights (y) and weights (x) of 100 coes
students are given:
Ex = 15000, Er = 2272500, Ly = 6800, Ey² =463025 and Exy = 1022250.
Find the correlation coefficient between height and weight.
 1T24.0b

Example 8. (i)Calculate coefficient of correlation from the following results:

n= 10, EX =100, EY= 150, EX 10)2= 180, E(Y- 15)²= 215, E(X-
10)(Ý - 15) = 60.
(ii)Calculate Karl Pearson's coeficient of correlation between X and Y
information: for the following
n= 12, X= 120, EY =130, E(X -8) = 150, E(Y- 10) = 200 and
EX-8)(Y - 10)= 50
Sol. (i) Mean of firstseries, X=. EX
2X_100 = 10
10

Mean of second series, y EY 150

= 15
n 10

Now, E(X - X)(Y Y)

EX-X)'.z(Y - Y)²
E(X -10)(Y - 15) 60
= 0.305
JE(X- 10) .E(Y- 15) J180 x 215
STATISTICALTECHNIQUES - | 267

Sx = X(X 8) =EX E8 = 120- (8 x 12)= 24

(iü)
Sy =2(Y - 10)= EY E10= 130 - (10 x 12)= 10
Jxy = EX - 8)(Y- 10)= 50 (given)
z2= E(X- 8)2 = 150
Ey² = (Y 10)= 200
n Exy - ExEy (12 x 5O) - (24 x 10)
Now,
JnEx2 -(Ex)² n Ey.- (Ey)2 J(12x 150) (24) J12x 200) (10)
360
= 0.2146.
J1224 J2300
 ffy 3fay- table tCorvela
on
h
way too preernfd and
a Trequeneyu Bivaviat
Dietribution
eorvelafion coiciert
af calculator
of
 xy Ulle ueW Varlabies u, Vare properly chosen.
Example 9. The following table gives according toage the frequency of marks obtained
by 100students in an intelligence test:
Age (in years)
18 19 20 21 Total
Marks
10-20 4 2 8
20-30 5 4 6 4 19
30-40 6 8 10 11 35
40-50 4 4 6 8 22
50-60 2 4 4 10
60-70 2 6

Total 19 22 31 28 100

Calculate the co-efficient of correlation between age and intelligence.

268 ATEXTBOOK OF
ENGINEERING
Let age and intelligence be denoted by x and y respectively. MATHEMATICS
Mid
value 18 19 20 21 f fu fu2
fuy
o8 -3
15 10-20 4 2 2 24 72
30
25 20-30 5 4 6 4 19 -2 -38 76 20
35 30-40 6 10 11 35 -1 -35 35
9
45 40-50 4 4 8 22 0 0
55 50-60 2 4 4 10 1 10 10 2
65 60-70 .2 3 1 6 2 12 24 -2
19 22 31 28 100 Total - 75 217 59
-2 -1 0 1 Total

fu - 38 -22 0 28 - 32

fu? 76 22 28 126

fuv 56 16 - 13 59

y- 45
, V = - 20
Let us define twonew variables u and vas u = 10

1
Efuv- Efu Efu
1 1
n n

1
59 (-75/- 32) 59 244
100 = 0.25.
1 643 X 2894
217 - (-75) | |126 -(- 32) 25
100 100 V4
 Rank Correlaton
65Di
n()
&pearrnans oroula or Rank
Nok Thà uu catled Sy; =0 alway!
eorYelaton. 4:) = Sx;-
Sd: =S(x;-
 erves as a check on calculations.

Example 10. Compute the rank correlation coefficient for the following data:
Person: A B C D E F G H J
Rank in Maths: 9 10 6 5 7 2 4 1 3
Rank in Physics: 2 3 4 5 6 7 9 10
270 ATEXTBOOK OF ENGINEERING M

Sol. Here the ranks are given and n = 10. MATHEMATIOS

Person R, R, D =R,-R,
A 9 1 8
64
B 10 2
64
6 3
D 4 1
1
7 5 2 4
F 2 6 -4 16
G 4 7 -3 9
H 8
I 1 9 -8 64
J 3 10 -7 49

ZD²=280
6ED2 6x 280
r=1 =1-1.697 =-0.697.
n(n? 10(100 1)|
Example 11, The marks secured by recruits in the selection test (X) and in the prof
ciency test (Y)are given below:
Serial No.: 1 2 3 4 5 6 7 9
X: 10 15 12 17 13 16 24 14 22
Y: 30 42 45 46 33 34 40 35 39
Calculate the rank correlation co-efficient.
Sol. Here the marks are given. Therefore, first of all, write down ranks. In each series.
the item with the largest size is ranked 1, next largest 2 and so on. Here n =9.
X 10 15 12 17 13 16 24 14 22 Total
30 42 45 46 33 34 40 35 39
Ranks in X (%) 9 5 8 3 7 4 1 6 2
Ranks in Y () 3 2 1 8 7 4 6 5

D=x-y 2 6 -1 -3 -3 -3 0

D² 4 36 4 1 9 9 9 72

62D² 6x 72)
r=l
n(n2 1)] |9x80| =1-.6 = 0.4
studentsin
Example 12. Rank correlation co-efficient of marks obtained by 10 diference in
Mathematics and English was found to be 0.5. It was later discovered that the Find
ranks in two subjects obtained by one of the students was wrongly taken as 3instead of?7
thecorrect rank correlation co-efficient.
eTÁTISTICAL TECHNIQUES- | 271

0.5, n = 10
Sol. Incorrect r; =

Incorrect ED2
nln - 1)(1 - r) 10 × 99 x 0.5
= 82.5
6 6
Now, correct D2 Incorrect ED2 -(3) 4+ (7)2 = 82.5 -9 + 49 = 122.5

62D2 6× 122.5
Correctr, =1 = 0.2575.
nín? - 1) 10(100 1)|
Rxample 13. Ten competitors ina beauty contest were ranked by three judges in the
ollowingorders:
First Judge: 6 5 10 3 2 4 9 7

Second Judge: 3 5 4 7 10 2 1 6 9

Third Judge: 6 4 8 2 3 10 5 7

Use the method of rank correlation to determine which pair of judges has the nearest
approach to common taste in beauty?
Sol. Let R, Rg, R, be the ranks given by three judges.
Calculation of rank correlation coefficient

Competitor R, Rg D12 DË3 D23 Dip Di D

=R,-R2 = R,- R, =R
A 1 3 6 - 2 -5 -3 4 25 9
B 6 5 4 1 1 4 1
5 8 9 -3 -4 -1 16 1
D 10 6 2 -4 36 4 16
E 3 7 1 =4 2 6 16 4 36
2 10 2 -8 8 64 64
G 4 2 1 -1 4 1 1
10 -1 -9 64 1 81
7 6 5 1 2 1 1 4 1
7 -1 1 2 1 1 4
n= 10
D,, ED3 ED
= 200 = 60 = 214

Kank correlation coefficient between first and second judges,

2
62D12
n(n2 1
6x 200 =-0.212
- 1)| 10 (99)
Kank correlation coefficient between first and third judges,
62D.13 2 6x 60
=1 = 0.636
n(n² - 1)| 10 x 99|
 ATEXTBOOK OF
272

Rank correlation coefficient between

second and third jud°es ENGINEERING MATHEU
2 |6x 214
62D23 = 1 10 x 99
=-0.297
n(n?- 1)
second judges is negative i.e., their
Correlation between first and
beauty test are opposite
to each other. Similarly, opinions of second and
opinions of first and third judges are of:
opintihoinsrd jureiggas
opposite to each other, but
correlation is positive. It
the
means similar
that their likings and dislikings are very much type a

41 TIED RANKS
 Reqreesion
Regre ion cooficienk
by by

bg bay bsy-(sy)2
grenior
byaCa-)

by (4-9)
betoean
Properlsa Coeticiert is the georetrie
) Correlation
the regenion Coeticienk
 ohe o th Togremion coiient. is gear
unity, the othe mut be les thas unib
that

byn by
6
ket byn > then, I-0
by»
byw bay
bay bya
u i4 bgu >l, tho
thar bya
3) Aritboeie eas
reais tha the Conatation coesficiest
prove that
le bave to
byu t bay
tor)
 Raqreon coticietbi an indapendlant ot
the orgb but bot os 3cale .
dot V Y-6
h
byn 6H:r koykfro bve
hou h (
y buy
6) The Corre laton eoesticient aol the to
ogrnion Cofficienk bave same Syn
y
byn

Sbee oL ahd oy Qe bot posifive,

by ahd Y have sare digh
bya ad, bay
stol,
Beusen tuoo lines o Rogvestios
Angle
abgle becfueen the too
o u the oeute
D Case t tuoo variat
eqresson linesI» the
&hoo tbat
abd y,
2
heis glopas
and

()
2

Qre positive

tas l
 Coeficien& are Og ard os 3)
4 fhe ragreson eoeticient b lorrelaton)
wbat would be the valu ot

Civeh, bya buy

bybay

4 Hobce, r Vo"t6 =

@2 caleutat the bnsay ragroscob toefficient yrom

the Qotoing:
3 5 1 6 17

lipear regenion eoots ae giveb by

byx bsxy - 2u2y abd bay DSy- Sniy
b S ( Sx)d nSy'-(sy

3 9 3 by a976

9
bay
10
6
36 40
S76
36 I768 S99
 Example 3. The following table gives age (x) in years of cars and annual maintenance
hundred rupees:
cOSt(y)in
X: 1 3 5 7 9
Y: 15 18 21 23 '22
Estimate the maintenance cost for a 4 year-old car after finding the regression equation.
Sol.
y

15 15 1
3 18 54
5 21 105 25
23 161 49
9 22 198 81

Lx = 25 Ey = 99 Xxy = 533 Ex2 = 165

Here, n=5

}*
Ex 25
x== = 5, y= Ey 99 = 19.8
n 5 n 5
n Excy - Ex Ey (5 x 533) (25x 99)
by =0.95
y*
n Ex-(Ex) (5x 165) (25)
Regression line of y on x is given by
y- y=b (x-x)
282 ATEXTBOOK OF ENGINEERING
MATHEMATICS
y-19.8 = 0.95 (x- 5)
y= 0.95x + 15.05
When x =4 years, y= (0.95 x 4) + 15.05 = 18.85 hundred rupees =1885.
ExamnoA .1l,, roonrd of n anlyis of a onol
 G en a paifiatty cutcyodd laboratry reco
a

e a analyiu ot a cornlataoo data, thi Jottouoing

Tesula only Qre
Vaiabe of x:9; Regvesios dquationg: &x-log +6620,
Ho- (8 y Qj4. hbat wese
ot
la) the mean Value of and y correlatin
the coefficient of
betoen abl

Ca) &ine botk the kires of rge%im pas throgh

the point (3, 9) .
koe ha

50y330

-3ay = -544

put the value o in O,

=13 abd y l
Hebe,
 (31
Witteh as

-13 =b5l9- 1)

(3) and
and A= 0"H54+535

Che agrehion

LO
oaltiping O,
Sope ineratr

rom , Poid slopa )

g-ge -
mN
 The
antb, 2= C* td, shou tbat
rspeclavel ý:
be+ od abd adtb
Oy- a E-al

Sola The rgreton lia

The egreuion lhe of

6
bay ad
bay
byu

buy

the
Linas pan through
sibce botk the rogye3 eco
pornt (a ),
±art6 and

ai-9= --b ard

dolring and ,
an- y

Oxa an- dcq ad Qd+b

(ac-)9 E-b-ad -aC
 xc ac- ey--be
d
(ac-) a =-be-d
- bc+ d
1-ac /
@6 o two rbdoro variables, abd y witk the
Jatne mean, the two gre91ion 0quaions are
y= ax+ b anot x dytB &hou that
g=
i d alo the Conon meab.

6olb Hore by ai bay

thes regeHion lies an
are
be o,
Let the Corppor mean and -m= (y-m)
y-o= a (n-m)

nd () oi th tbe gives aghe

eoroparirg oCt- a), B= otl-
)
b=
l-a

hegreston lirnes pan thougk e, ).

siee wi hold
abd y a+ b

b-p
ola-a): b-ß
b-ß
m=b-B
d-a
 data given bela..
Example 7. (i)Obtain the line of regression ot y on x for the
2.60 2.95 3.42
1.53 1.78
40.00 45.80 53.50.
y: 33.50 36.30
of
(iü) The following data regarding the heights (y) and weights (x) F100
are given:
college studer
Lx= 2272500, £y = 6800, Ey² =463025 and Exy =
Ex = 15000,
1022250.
Pind the equation of regression line of height on weight. (A.K.T.U. 207
Sol. (i) The line of regression ofy on x is given by
y- ù=b, (x-)
where yx
b. is the coefficient of regression given by
n Exy - Ex Ey
n Ex -(Ex)
Now we form the table as,

1.53 33.50 2.3409 51.255

1.78 36.30 2.1684 64.614
2.60 40.00 6.76 104
2.95 45.80 8.7025 135.11
3.42 53.50 11.6964 182.97
Ex= 12.28 2y = 209.1 Ex =32.6682 }xy= 537.949
Here, n =5

(5x 537.949) (12.28 x 209.1)

(5x 32.6682) - (12.28) =9.726

Also, 12.28 209.1

and y = Ly
mean X= = 2.456
n 5 =41.82
n 5
From (1), we get
y-41.82 = 9.726(% 2.456) = 9.726x 23.887
y= 17.932 + 9.726x
(iü) }x 15000
n = 150, y= 2y 6800 = 68
100 n 100
Regression coefficient of yon x.
n Zxy - Zax Ey
yx (100 x 1022250) -(15000 x 6800) = 0.1
nEx -(Ex)
(100 x 2272500) -(15000)*
 TEECHNICQUES-| 285
TISTICAL
STA

Regression line offheight (y) on weight (x) is given by

y- ù=b, (x-)
y- 68 = 0.1(* - 150)
y= 0.1x + 53.
Example 8. For 10 observations on price (x) and supply (y), the following data were
appropriate units):
ained(in
Lx = 130, Ey = 220, Ex' = 2288, Ly².= 5506 and Exy = 3467
Obtain thetwo lines of regression and estimate the supply when the price is 16 units.

Sol. Here, n= 10, x= = 13 and y= Ey = 22

Regression coefficient of y on x is
n Exy - Ex Ey (10 x 3467) - (130 x 220) =1.015
b= n Ex-(Lx)? (10 x 2288) - (130)
Regression line ofy on x is
y-ù=b(x-)
y- 22 = 1.015(* 13)
y= 1.015x +8.805 .(1)
Regression coefficient of x on y is
n Exy - E Ey (10× 3467) - (130 × 220) = 0.9114
b., =
nEy² - (2y? (10 x 5506) - (220)
Regression line of x on y is
x-=b(y -9)
x- 13 = 0.9114(y - 22)
x =0.9114y 7.0508 ...(2)
Since we are to estimate supply (y) when price (x) is given therefore we are to use
regression line ofy on x here.
When x =16 units,
y= 1.015(16) + 8.805 = 25.045 units.
Example 9. The following results were obtained from records of age (x) and systolic
blood pressure (y) ofa group of 10 men:

Mean 53 142
Variance 130
and (x - )(y - y) = 1220
Findthe appropriate regression equation and use it to estimate the blood pressure of a
man whose age is
45.
Sol. Given
Mean I =53 and y = 142; Variance o= 130
n= 10 and E(x-Ty- y) =1220
 ATEXTBOOK OF
286
ENGINEERING
Since we are to estimate blood pressure (y) of a 45 years old man, we will fnd MATHEMATIOS
line of y on x.
Regression coefficient regres ior
oy =
Z(x-)(y-9)y
nG, Oy
E(x-) (y-ù) 1220
= 0.93846
n o, (10) (130)
Regression line of y on x is given by
y-ù = b, (x-)
y-142 = 0.93846 (x 53)
y= 0.93846 x+ 92,26162
When * = 45, y= 134.49
Hence the required blood pressure = 134.49.
Example 10. The following results were obtained from marks in Applied Mechonis
and Engineering Mathematics in an examination:
Applied Mechanics (x) Engineering Mathematics (y)
Mean 47.5 39.5
Standard Deviation 16.8 10.8
r=0.95.

Find both the regression equations. Also estimate the value of yfor x =30.
Sol. Here, =47.5, y =39.5
o, =16.8, 9, = 10.8 and r=0.95.
Regression coeficients are
10.8
b, =r 0.95 x
16.8
=0.6107

and 16.8
b =r =0.95 x = 1.477.
Oy 10.8
Regression line of y on x is
y-ù=br-)
y- 39.5 = 0.6107 (% -47.5) = 0.6107x
29.008
y= 0.6107x + 10.49
Regression line ofx on yis
x-=b, (y- )
x-47.5 = 1.477 (y - 39.5)
x = 1.477y 10.8415
Puttingx = 30 in equation (1), we
get
y=(0.6107)(30) + 10.49 = 28.81.
STATISTICALTECHNIQUES 287

Examplel1. The equations of two regression lines, obtained in .correlation

a analysis of
ANDbserationsare:
5x =6y +24 and 1000y = 768x - 3608.
What isthe correlation coefficient? Showthat the ratio of coeficient of variability of x to
5 (A.K.T.U. 2019)
hatofyis24
What is the ratioof variances of xand y?
Sol. Regression line of xony is
5x = 6y + 24
6 24
5 5
6
...(1)
by5
Regression line of yon x is
1000y = 768x -3608
y=0.768x - 3.608
b,=0.768 ...(2)
6
...(3)
From (1), 5

From (2), =0.768 ...(4)

Multiplying equations (3) and (4), we get

p2= 0.9216 ra0.96 ...(5)
Dividing (4) by (3), we get
2
6
= 1.5625.
2 5x0,768
Taking square root, we get
5
O125 =4 ...(6)

Since the regression lines pass through the point (, y),we have
5~ =6y + 24
1000 ù=768T 3608.
Solving the above equations for and ù, we get =6, ù =1.
Coefficient of variability of *=

Coefficient of variability of y=
1 5 5
Required ratio = =

6 4 24 |Using (6)
288 A TEXTBOOK OF ENGINEERING

Example 12. Apanel of tw0 judges, Aand B, graded seven TV serial

MATHEMATICSby
awarding marks independently as shown in the following table:
performances
3 4
Performance 2 7

Marks by A 46 42 44 40 43 41
45
Marks by B 40 38 36 35 39 37
41
The eighth TV performance which judge Bcould not attend, was awarded 37
marksbo by
judge A. If the judge B had also been present, how many marks would be expected to haye
awarded by him to the eighth TV performance?
Use regression analysis to answer this question.
Sol. Let the marks awarded by judge A be denoted by x and the marks awarded h.
judge B be denoted by y.
Zx 46 +42 + 44+ 40 + 43 + 41+ 45
Here, n=7; X= =43
n 7

Ey 40 + 38 +36+ 35 +39 +37+ 41

y= =38
n 7

Let us form the table as

46 40 1840 2116
42 38 1596 1764
44 36 1584 1936
40 35 1400 1600
43 39 1677 1849
41 37 1517 1681
45 41 1845 2025
x = 301 Ey = 266 Zxy = 11459 Ex? = 12971

Regression coefficient,
n Exy - Ear Ey (7x 11459) (301 x266)
Oyx =0.75
nEx -(Ex) (7x12971) - (301)
Regression line ofy on xis given by
y- ù=b,,(*)
y- 38 = 0.75(% - 43)
y= 0.75x +5.75
When x=37, y=0.75(37) + 5.75 = 33.5 marks
Hence, ifjudge B had also been present, 33.5 have beer
markS would expectedto
awarded to the eighth TV be
performance.
 -|
TICAL TECHNIQUES 289
STATIS
Example 13. Two variables x and y have zero means, the same variance o and zero
correlation, showthat:
U =X COS +y Sin oa and V=X Sn a-y cos 0.
variance o and zero correlation. (U.P.T.U. 2007)
same
havethe
Sol. We are given that
r (x,y) = 0 ’ Cov (x, y)= 0, o² = o = o
We have, 2= V(x cos .+y Sin 0)
= cos² a V (x)+ sin a V(y) +2 sin a cos a Cov (%, y)
= (cos a+ sin a) o² |:: Cov (x, y) = 0

Similarly,
Cov (u, v)= E [(u - k) (u - ))
=E ( cOs +y sin a- x c0s - y sin a)
a)]
(* Sin a-y cos - x sin a + y cos
= E[(* cos a +y sin a) (x sin a-y cos a)] |:: I =0= y
= E(«) E(y)]sin a cos a + E(y) (sin² a - cos a)
=0 |: o'= g,?= o² and E(uy) = 0
Cov (u, v) = 0.