,.

MATHEMATICAL
4.12 FouNDA.r10
Ns Fo~ co~
IATION
4.5. LIMITATIONS OF MEAN DEV Pui
~
t~ ,,
e cen tral tendenc .
The me dia n is not a rep res ent ativ d · · d'ase r·Ies in
. b'l 't Th us, th e me an ev1ation abo ut Yin c ~
degree of van a 1 1 y. ~~
alcul ated of a1
. o, the sum of the dev iafions &om ~
can not be full y reli ed upo n. Als fd · · the Ill for suci
· h o ev1 atio ns from the median S
ignored) 1s mo re tha n t e sum · 0, the"' dean. (' t,
. . •uean
· t'fi Th ere1~ore, m · ma ny cases mean devia ev1atj
mea n is not ver y sc1en 1 1c. . . . ' tion . u"
gives ~
.
me an dev iati on 1s mad e on the b .
results. Also, the calc ula tion of . as1s of ab
. solutevai ,,
48%,i
not be giv en any furt her algebr aic treat
deviations and as suc h the y can . . In
Ue1,
lllent. ~-
f d 1sp ers1on to overcome the problem
req uire some oth er me asu re o g rts~ch cases a'
ch sha ll be stud yin g in the followin. a icle. 11 i
deviation is found ver y use ful whi
.)
4.6. STANDARD DEVIATION (S.D
tive
isti cal dat a is defined as the posi 8
The sta nda rd dev iati on of a stat metic m;::rero,1 ,
dev iati ons of item s from the arith
arit hme tic mea n of the squ are of d as sigma). It is also knoft~1
und er considerati• on. It is gen era lly den ote d by cr (rea own a,
•
mean squ are dev1at1on.
n
.I; <xi-x}2
= 1/-i_=_I__ __
For an ind ivid ual series: S.D. (cr) n
mean.
of var iab le and xis the arithmetic
where Xi, x 2 , ... xn are the valu es
tten as
The above formula may also be wri
S.D. (cr) = y---;;-/ - x-2
lr.x
r- l:(x , -x) 2
1:(x / -2x -x +x
n
I 2
) rx/
_ _ _ ,X
2rx; -.
n n
rr2 ·
LX·2 -2 - 2 _ _:..!.- -1
= - -1--2 x +x - n
n
n 2
1: t<xi -x>
. = 1/_i_=_I_-: -:-- ~
For a frequency distribution : S.D
(cr)
N
. deJ1 t,<10f'
i n) and N = l: f;.
whe re f; is the freq uen cy of x, (I
$ $
0
. ce and is
iati on is call ed var ian
;i'~
stan dar d dev
VARIANCE. The square of tiu, . en ill -gblt
a e are grv es oftbe vefl
tion , whe n the valu es of the var1. bl
Remark, In a freq uen cy dist ribu are tak en as the va1u
mid -po ints
of clas ses, then thei r resp ecti ve
..
 orsPERSION 4.13
5u1<ES oF
116/1
I sro CALCULATE THE STAN DARD DEVIATION
r.4efHOD
41,
' t~etb od:
pifeC
4,1,1· "£{;,(xi -xi ... (1)
(J = N
We)tnOW that
"£{;, (x/ - 2xix + x2)
N
1 f,· X · 2 -
-"£ 2- 1 ~ f,
X-,:., 1· X·1
1 f, ·
-2 .-"£
+X 1
N
1 1
N N
~ f, - 2-x. x- + x-2 [·: x = "£~xi and N = "£{;,]
JN1 ,:., i xi
2
=
= J!1:t;,x/-x2
cr= .IT(.I:-T]2.I: -l'.,x-.2 -1'.x. ... (2)
4.7,2. Change of Origin (Shor t-cut Meth od)
~ ~en _the mean of distrib ution is a
decim al ~ac~i on, the ~alcu lation work becomes
arbitr ary numb er A, and find
!hed : ~ m such cases we take assum ed mean , which 1s any
0
eVJation from this mean.
ary numb er A i.e., A is the new
or~n~t di == xi - A i.e., di are devia tions of xi from an arbitr
d == x -A ·: di= xi -A ⇒ r, {;,di= r, f;,x~- ~r, {;,
⇒ L ft di r, {;, xi _ A I,{;, ⇒ d x - A
= =

l
r N N N
d.- d - d (- - - . .. (3)
! - i - X - A) = X !. - A - X + A = X I. - X
Now
cr 2 1
X

_
- N L t;(xi - x)2
i!_ - 1 ~
- N~ t; (di - d)2
.. .(4)
cr/ ::: cr/ : : :, crx :::: crd
 MATH EMAT ICAL FO -......
UNoATroNs ,
4.14
c0~Plt
~OR
Using (2), equat ion (4) can be writt en as t(
C1x = Ef,d,' -(Et;/')'
Result (4) show s that standa rd devia tion rema ins un
changed b
Y the chang
ula takes the form eor~,,
Remu k : For indiv idual serie s, the above form
- _( r:,J 2
2
r.,d
S.D. = .!-f
4. 7.8. Chan ge of Orig in and Chan ge of Scal e
(Step Devi ation Method)
X· - A
Let u I. = - 1-h - , wher e A and hare arbit rary const ants.
x I. - A=h uI. ⇒ x I. =A+ hu I.
⇒ x = A+h . u ⇒ xi - x =x i - (A + h . ii )
⇒ x /.. - x = A+ hu . - I
(A+ hu) ri~,,_,i
⇒ x; - x = h(ui - u)
Now cr2 -_
x
"
N1 L..J //x; - x)2
1
= N L I; [h (u; - u )]2 [l's.·;
= h2 __!_ " {·: h ist,....
N L..J f; (u,- - u )2
= h 2 cr 2
u
ax= h. /t •~
l
-
f /; u,2 - ( f NI; u,12 I • J
where f, is the freque ncy of x, ( l s i s n /, a nd u, ~~ , -
stJll ~-
x. - A ( l ~i < fl ): a an d hare cun
Thus ~
s.n. (o) h ~ - ( ~ ~
-- -- --2
) of;di
Equat ion (BJ shows that S .D. is indePN<lnde nt ofNchang
e or origin but not cbaage
 N 4.15
rsrERSIO
~ie:
~
l
. al prob ems
h may be taken as the width of class-intervals.
fri 11umtric
J.
SD of a set of n 1 observations are X1 and s1 respectively, while the mean and
and · ·
Jfthe mean fn observations are X2 and s2 respectively, then the S.D. of the combined set
2, other set o 2
s.D. of an observations is given by
0
f(/1 + ri2)
1 I,-------=---------=-
2
n1 SJ + n2 S2
SD=--~--=-~+
2
n1 n2 (x1 - X2)
~
2
. . • n1 + n2 (n1 + n2)
- .... ··---····...;•+;;~;:x.::;~i - ·,· ' . .,. '""' ,"~~,;,{:-
~►- .. _.,. . •..•A
_•~•-·-··
,t,~,••••••••
,..... ·~,~-:.,__ ·,,.,.:..
•" .,. .-..~ ""'$OLVED
, .£XAMPl:ES·~_:,,f0r
L, ,,. . . . ., , ~.r
-~,
___ -·
, .. - .. \., ~ •••••~••~,-£.,....i&
7;-,,....,~~ • - ..-.:~- <;rw~ ~
,_;."d?Z•f'-·•
,J"fe,., 4".;:
. ..,,, • •-:,ik'
.. ..,_,{ __~ , ~ ,_
' ..., ;
- . . .....
,,...,,,
.••
Example 1. Find the standard deviation ofll, 14, 15, 17, 18.
soiution. We may first prepare the following table:
,--S.No. xi (xi - x) (xi- x)2
(x = 15)
1 11 -4 16
2 14 -1 1
3 15 0 0
4 17 2 4
5 18 3 9
n=5 fai = 75 :E(xi - x) 2 = 30
-
Here - ~- 75
X::-1 = - = 15
n 5
~ow S.D. = Jr(x,n- X)'
/30
= vs
= ✓B = 2.45_
Example 2
· The scores of batsman A in 10 innings were
(~ ~il!d t~ V . 38, 70, 48, 34, 42, 55, 63, 46, 54, 44.
(4) W1io ariance of the scores.
So1 t would b th . .
11tlon (' e e variance if he had scored 10 runs more in each innings.
~~ . i) We Illa fi
~ 48 Y rst prepare the following table :
~
•~
,. L......,~\
4.16 MATHEMATICAL FOVNo/\TroNs F
0~ Co/lfplJ
I -
di= xi -A ---,- ___ _ " ' r11~:
x. i
(A =48) I ..,, ,~ l ~'
r
I l
38
70
-10
22
~
484
48 0
0
34 -14 196
42 -6 36
55 7 49
63 15 225
46 -2 4
54 6 36
44 -4 16
-
LXi = 494 I: di = 14 I
l:d/ = 1146
~
on. So we shall be using short cutme~
Here the mean comes out to be a decimal fracti
2
I:d- - (I:d
= --;:-- ~) -/ [·: Variance =(S.D
Variance (cr2)
1146 -
= 10
10
(14) 2
= 114·6 - 1-96 = 112·64
(ii ) If the batsm an had score d 10 runs
more in each innings, the new set of scores!
48, 80, 58, 44, 52, 65, 73, 56, 64,54
Calc ulati on of new varia nce : Let A = 58
dI.2 -
-
xi d I. =x.t -A
(A =58)
100
48 -10
80 484
22
58 0
0
44 -14 196
52 -6 36
65 7 49
73 15 225
56 -2 4
64 6 36
54 -4 16--
:Exi = 594 Y.d; == 14 x.dl 2 = 1146
j
 4.17
~
,,,RES~
Variance I!/ -(I:i
= r
2
= 1146 -(14)
10 10
= 114-6-1-96 = 112-64
the scores are increased by 10 in each innings, the variance still remains
fbus when
unchanged, Find the variance and standard deviation for the following data :
Example 3.
[ _
I_;_/'/- ;=---r1------::-;-;1-rl------::;;;7----,1---:::-::o- rl--:::4-:--'l--3=-=::---,1
Solution. Calculation of variance :
- x (xi - x)2 x)2
\ ~ f;,xi xi- f;,(xi -
4 3 12 -10 100 300
8 5 40 -6 36 180
11 9 99 -3 9 81
17 5 85 3 9 45
20 4 80 6 36 144
24 3 72 10 100 300
32 I 32 18 324 324
~
N=30 L~Xi =420 "'if/xi- x) 2 = 1374
- "'if X; - 420 = 14 [·: N= 30]
X = ~ - 30
tJj
Variance (a 2) = _!_
N LJ
l

'°'
7
=I
f,-(x .
l l
-xi = __!_
30
x 1374 = 45-8
S.D. (a) = ✓45.8 = 6-77.
1;11 ~Ple4 p·
"V-![ • ln,d th . . . . . .
· e mean and standard deviation for the following data using sho, t-cut
62 63 64 65 66 67 68
11llio 12 29 25 12 10 4 5
n. Let A."' 64
➔ rn
MATHEM ATICA L Fo
~yf u
NDAT' toN
S PoR
. Co'-fPu
. n.
an d sta nd ar d devlatio . r,,
Calcu lat,on of mean
,.-- - , , - -/.-,- , , -d,.-=-x,. ;- -- , ~;,
--6-4 ~ ,-
x,
2 -4
60
1 -3
61
12 -2 4
62
29 -1 1
63
0 0
64 25
1 1 12
65 12
2 4 20 4()
66 10
3 9 12 36
67 4
4 16 20 80
68 5
N= lO O
I,(; d - 0 64
Mean (x) = A + _ I_ , = 64 + -100 =
N
-
S.D . = , / T, tdi2 -e : id)2
~
~ = -v2
-86 = 1-69 (ap pro x).
=
0
ngdisfTWIJ•
d deviation for the followi
e the mean , variance an d sta nd ar
Ex am ple 5. Ca/,eu/,at -- 90- 10!
60 - 70 70 -8 0
30 - 40 / 40 -5 0 50 -6 0
Class : 3
15 8
3 7 12
Fr eq ue nc y:
e an d S.D . :
So lu tio n. Ca lcu lat
ion of mean, va ria nc
I Mid-valu e Freque nc y / !; x,
CLa c;s
x, t;
11J;J
35 3 105
10 - 40 ~
40 !j (J 45 7 315
135
fi () (j() fjfj 12 66 0
13Si
60 - 70
n,.,> 1/5 975
75 8 (j()()
70 - 80
80 - 90 85 :1 25 5
90 - 100 95 2 190
N = 50 I f, x,
=3 10 0
 N_ _ _ _ _ __
ER_sr_o_
oF rsr_
5 0
~ --
Mean (x ) = Zt'ix
N
i -_ 310 0 =62
50
7
.I t'i Cxi - x>2
1
t =l
Variance (cr2) = N = 50 X 10050 = 201
S.D. (cr) = .J201 =14-17
and
ribution ·
Find the standard deviation of the following dist ·
ExamPie 6·
-
Marks:
No. of students :

10- 20 20- 30 30 - 40 40- 50 50 - 60 60-

5 12 15
70 70 -80
20 10 4 2
Solution. Let A = 45 and h = 10
Wemay now prepare the following tabl e :
X· - A
Marks Mid-values No. of U,=
l
l
h t: u.
l l
{; u .2
l l
(x.)
I
students (f) (A =45; h = 10)
-3 -15 45
10 - 20 15 5
20 - 30 -2 - 24 48
25 12
30 - 40 -1 - 15 15
35 15
40 - 50 0 0
45 20 0
50-60 10 10
55 10 1
60 -70 8 16
65 4 2
~ 18
........__-_
75 2 3 6
N : 68 1:(;t u l. = -80 r ft u.2 =152
l
Mow
S.D.(a) = h . . p:·{; u,2 -( L ~u.y =10 x J16~ - ( -6: o)'
~~arr, Pie 7, 'I'he . = 10 x .J2•24 - O-19 = 10 x 1-43 = 14·3 (nearly).
· n bel ow :
· les (in m m) dra wn in a design are give
1l,D~'heto•• ,. diameters 0f circ