Measures of Dispersion: Quantitative Aptitude & Business Statistics

Measures of Dispersion
Quantitative Aptitude & Business Statistics

Quantitative Aptitude & Business Statistics:
Measures of Dispersion 2
Why Study Dispersion?
An aver age, suc h as t he mean
or t he medi an onl y l oc at es t he
c ent r e of t he dat a.
An aver age does not t el l us
anyt hi ng about t he spr ead of
t he dat a.
What is Dispersion
Di sper si on ( al so k now n as
Sc at t er ,spr ead or var i at i on )
measur es t he i t ems var y f r om
some c ent r al val ue .
I t measur es t he degr ee of
var i at i on.
Si gni f i c anc e of Measur i ng
Di sper si on
To det er mi ne t he r el i abi l i t y of
an aver age.
To f ac i l i t at e c ompar i son.
To f ac i l i t at e c ont r ol .
To f ac i l i t at e t he use of ot her
st at i st i c al measur es.
Pr oper t i es of Good Measur e of
Di sper si on
Si mpl e t o under st and and easy
t o c al c ul at e
Ri gi dl y def i ned
Based on al l i t ems
A meanabl e t o al gebr ai c
t r eat ment
Sampl i ng st abi l i t y
Not undul y af f ec t ed by Ex t r eme
i t ems.
Absol ut e Measur e
of Di sper si on

Based on sel ec t ed
i t ems
Based on al l i t ems
1.Range
2.I nt er Quar t i l e Range
1.Mean Devi at i on
2.St andar d Devi at i on

Relative measures of Dispersion
Based on
Selected
items
Based on
all items
1.Coefficient of
Range
2.Coefficient of
QD
1.Coefficient of MD
2.Coefficient of SD &
Coefficient of Variation
A smal l val ue f or a measur e of
di sper si on i ndi c at es t hat t he dat a
ar e c l ust er ed c l osel y (t he mean i s
t her ef or e r epr esent at i ve of t he dat a).
A l ar ge measur e of di sper si on
i ndi c at es t hat t he mean i s not
r el i abl e (i t i s not r epr esent at i ve of
t he dat a).

The Range
The si mpl est measur e of
di sper si on i s t he r ange.
For ungr ouped dat a, t he
r ange i s t he di f f er enc e
bet w een t he hi ghest and
l ow est val ues i n a set of
dat a.
RANGE = Hi ghest Val ue -
Low est Val ue

The r ange onl y t ak es i nt o
ac c ount t he most ex t r eme
val ues.
Thi s may not be
r epr esent at i ve of t he
popul at i on.

The Range Example
A sampl e of f i ve ac c ount i ng
gr aduat es r eveal ed t he
f ol l ow i ng st ar t i ng sal ar i es:
22,000,28,000, 31,000,
23,000, 24,000.
The r ange i s 31,000 - 22,000
= 9,000.
Coefficient of Range
Coef f i c i ent of Range i s
c al c ul at ed as,
Coef f i c i ent of Range =
1698 . 0 =
+
S L
S L
Fr om t he f ol l ow i ng dat a
c al c ul at e Range and
Coef f i c i ent of Range
Marks 5 15 25 35 45 55
No .of
Students
10 20 30 50 40 30
Largest term (L)=55,
Smallest term (S)=5
Range=L-S=55-5=50
Coefficient of Range
833 . 0
5 55
5 55
S L
S L
=
+
=
+
=
From the following data ,calculate
Range and Coefficient of Range
Marks 0-10 10-20 20-30 30-40 40-50 50-60
No .of
Students
10 20 30 50 40 30
Low er l i mi t of l ow est c l ass
(S)=0
Upper l i mi t of hi ghest c l ass
(L)=60
Coef f i c i ent of Range
000 . 1
0 60
0 60
S L
S L
=
+
=
+
=
Mer i t s Of Range
1.I t s easy t o under st and and
easy t o c al c ul at e.
2.I t does not r equi r e any
spec i al k now l edge.
3.I t t ak es mi ni mum t i me t o
c al c ul at e t he val ue of Range.

Limitations of Range
I t does not t ak e i nt o
ac c ount of al l i t ems of
di st r i but i on.
Onl y t w o ex t r eme val ues ar e
t ak en i nt o c onsi der at i on.
I t i s af f ec t ed by ex t r eme
val ues.
Li mi t at i ons of Range
I t does not i ndi c at e t he
di r ec t i on of var i abi l i t y.
I t does not pr esent ver y
ac c ur at e pi c t ur e of t he
var i abi l i t y.

Uses of Range
I t f ac i l i t at es t o st udy qual i t y
c ont r ol .
I t f ac i l i t at es t o st udy var i at i ons
of pr i c es on shar es ,debent ur es,
bonds and agr i c ul t ur al
c ommodi t i es.
I t f ac i l i t at es w eat her f or ec ast s.
I nt er quar t i l e Range
The i nt er quar t i l e r ange i s
used t o over c ome t he
pr obl em of out l yi ng
obser vat i ons.
The i nt er quar t i l e r ange
measur es t he r ange of t he
mi ddl e (50%) val ues onl y
I nt er quar t i l e r ange = Q
3
Q
1

I t i s somet i mes r ef er r ed t o
as t he quar t i l e devi at i on or
t he semi -i nt er quar t i l e
r ange.

Ex er c i se

The number of c ompl ai nt s
r ec ei ved by t he manager of
a super mar k et w as
r ec or ded f or eac h of t he
l ast 10 w or k i ng days.
21, 15, 18, 5, 10, 17, 21,
19, 25 & 28

I nt er quar t i l e Range Ex ampl e

Sor t ed dat a
5, 10, 15, 17, 18, 19, 21,
21, 25 & 28

75 . 13 75 . 3 10
) 10 15 ( 75 . 0 2
75 . 2
4
11
4
1
1
1
1
= + =
+ =
=
=
+
=
item
Q
Q
N
Q
( )
23 ) 4 ( 25 . 0 21
) 21 25 ( 25 . 0 21
) 8 9 ( 25 . 0 8 25 . 8
4
33
4
1 3
3
3
3
= + =
+ =
+ = =
=
+
=
thitem thitem Q
Q
N
Q
I nt er quar t i l e r ange = 23 13.75
= 9.25
Co-ef f i c i ent of Quar t i l e
Devi at i on =
1979 . 0
1 3
1 3
=
+
=
Q Q
Q Q
From the following data ,calculate Inter
Quartile Range and Coefficient of
Quartile Deviation
Mar k s 0-
10
10-
20
20-
30
30-
40
40-
50
50-
60
No .of
St udent s

10

20

30

50

40

30
Calculation of Cumulative frequencies
Mar k s No. of
St udent s
Cumul at i ve
Fr equenc i es
0-10
10-20
20-30
30-40
40-50
50-60
10
20
30f
50
40 f
30
10
30 c f
60Q1 Cl ass
110 c f
150Q3Cl ass
180
180
L1
L3
Q
1
= Si ze of N/4
t h
i t em = Si ze of
180/4
t h
i t em = 45
t h
i t em
Ther e f or e Q
1
l i es i n t he Cl ass
20-30
C
f
f . c
4
N
L Q
1

+ =

25
10
30
15
20
10
30
30
4
180
20 Q
1
=
+ =

+ =
Q
3
= Si ze of 3.N/4
t h
i t em = Si ze of
3.180/4
t h
i t em = 135
t h
i t em
Ther e f or e Q
3
l i es i n t he Cl ass
40-50
C
f
f c
N
L Q

+ =
.
4
. 3
3
25 . 46
10
10
25
40
10
40
110
4
180
. 3
40 3 Q
=
+ =

+ =
I nt er Quar t i l e Range=Q
3
-Q
1
=46.25-25=21.25

Coef f i c i ent of Quar t i l e Devi at i on

1 3
1 3
Q Q
Q Q
+
=
=0.2982
Merits Of Quartile Deviation
I t s easy t o under st and and easy
t o c al c ul at e.
I t i s l east af f ec t ed by ex t r eme
val ues.
I t c an be used i n open-end
f r equenc y di st r i but i on.
Li mi t at i ons of Quar t i l e Devi at i on
I t i s not sui t ed t o al gebr ai c
t r eat ment
I t i s ver y muc h af f ec t ed by sampl i ng
f l uc t uat i ons
The met hod of Di sper si on i s not
based on al l t he i t ems of t he ser i es .
I t i gnor es t he 50% of t he
di st r i but i on.

Mean Deviation
The mean devi at i on t ak es i nt o
c onsi der at i on al l of t he val ues
Mean Devi at i on: The ar i t hmet i c
mean of t he absol ut e val ues of
t he devi at i ons f r om t he
ar i t hmet i c mean

n
x x
MD

=
Wher e: X = t he val ue of each
obser vat i on X = t he ar i t hmet i c
mean of t he val ues
n = t he number of obser vat i ons
|| = t he absol ut e val ue (t he si gns of
t he devi at i ons ar e di sr egar ded)
Mean Devi at i on Ex ampl e
The weights of a sample of crates
containing books for the bookstore are
(in kgs.) 103, 97, 101, 106, 103.
X = 510/5 = 102 kgs.
|x-x| = 1+5+1+4+1=12
MD = 12/5 = 2.4
Typically, the weights of the crates are
2.4 kgs. from the mean weight of 102 kgs.


I f t he dat a ar e i n t he f or m of a
f r equenc y di st r i but i on, t he mean
devi at i on c an be c al c ul at ed usi ng t he
f ol l ow i ng f or mul a:

Wher e f = t he f r equenc y of an
obser vat i on x
n = f = t he sum of t he f r equenc i es

Frequency Distribution Mean
Deviation

=
f
x x f
MD
_
| |
Fr equenc y Di st r i but i on MD
Ex ampl e

Number of
out st andi ng
ac c ount s

Fr equency

f x

|x -x |

f |x -x |

0 1 0 2 2
1 9 9 1 9
2 7 14 0 0
3 3 9 1 3
4 4 16 2 8
Tot al : 24 f x
= 48
f |x -x | =
22
=
f
fx
x
_

=
f
x x f
MD
_
| |
MD =22/24 =0.92
mean = 48/24 = 2
Calculate Mean Deviation and
Coefficient of Mean Deviation

Mar k s
(X)
No. of
St udent s
Cumul at i ve
Fr equenc i es
0-10
10-20
20-30
30-40
40-50
50-60

10
20
30
50 f
40
30
10
30
60 cf
110
150
180
180
L
Median
Class
Cal c ul at i on of Mean devi at i ons f r om
Medi an
2
N
2
180
=
=90
t h
i t em
Medi an i n Cl ass =30-40,L=30,c .f =60,
f =50 and c =10

36 6 30
10
50
30
2
180
30
c
f
f . c
2
N
L M
= + =

+ =

+ =
Cal c ul at i on of Mean Devi at i on of
Medi an
Mean Devi at i on f r om Medi an

N
M x f

=
333 . 11
180
2040
= =
Coef f i c i ent of Mean
Devi at i on

3148 . 0
36
333 . 11
Median
ion MeanDeviat
= =
=
Merits of Mean Deviation
I t i s easy t o under st and
I t i s based on al l i t ems of t he
ser i es
I t i s l ess af f ec t ed by ex t r eme
val ues
I t i s usef ul smal l sampl es w hen
no det ai l ed anal ysi s i s r equi r ed.
Li mi t at i ons of Mean Devi at i on
I t l ac k s pr oper t i es suc h t hat
(+) and(-)si gns w hi c h ar e not
t ak en i nt o c onsi der at i on.
I t i s not sui t abl e f or
mat hemat i c al t r eat ment .

I t may not gi ve ac c ur at e
r esul t s w hen t he degr ee of
var i abi l i t y i n a ser i es i s ver y
hi gh.

Standard Deviation
St andar d devi at i on i s t he most
c ommonl y used measur e of
di sper si on
Si mi l ar t o t he mean devi at i on,
t he st andar d devi at i on t ak es
i nt o ac c ount t he val ue of ever y
obser vat i on
The val ues of t he mean
devi at i on and t he st andar d
devi at i on shoul d be r el at i vel y
si mi l ar .

St andar d Devi at i on
The st andar d devi at i on uses t he
squar es of t he r esi dual s
St eps;
Fi nd t he sum of t he squar es of
t he r esi dual s
Fi nd t he mean
Then t ak e t he squar e r oot of
t he mean

N
x x

=
2
_

Fr om t he f ol l ow i ng dat a c al c ul at e
St andar d Devi at i on
5,15,25,35,45 and 55

N
X
X

=
30
6
180
X = =
078 . 17
6
1750
N
) X X (
2
= =
=

Fr equenc y Di st r i but i on SD
If the data are in the form of a frequency
distribution the standard deviation is given by
2
2
2
N
x . f
N
x . f
N
) X X .( f

From the following data ,calculate Standard
Deviation .
Marks
(X)
5 15 25 35 45 55
No. of
Students
(f)
10 20 30 50 40 30

Mean =
35
180
6300
N
x . f
X = = =

Mar k s
(X)
No. of
St udent s
(f )
(X-
35)=x
f x
2
5
15
25
35
45
55
10
20
30
50
40
30
-30
-20
-10
0
10
20

9000
8000
3000
0
4000
12000

N=180 f x
2
=36000
( )
2
2
2
2
X
N
x . f
N
x . f
N
x . f
=

=14.142
Properties of Standard Deviation

I ndependent of c hange of or i gi n
Not i ndependent of c hange of Sc al e.
Fi x ed Rel at i onshi p among measur es
of Di sper si on.
=
=
5
4
MD
3
2
QD
Thus SD is never less than QD and MD
In a normal distribution there is fixed relationship

3
2 1 +1 +2
+ 3
Bell -Shaped Curve showing the relationship between and .
68.27%
95.45%
99.73%
Minimum sum of Squares; The Sum of
Squares of Deviations of items in the
series from their arithmetic mean is
minimum.
Standard Deviation of n natural numbers
12
1 N
2
=
Combined standard deviation
2 1
2
2 2
2
1 1
2
2 2
2
1 1
12
N N
d N d N N N
+
+ + +
=
Where =Combined standard Deviation of
two groups
=Standard Deviation of first group
N1=No. of items of First group
N2=No. of items of Second group

= Standard deviation of Second group
12


12
1 1
X X d =
12
2 2
X X d =
Where is the combined mean
of two groups
12
X
Mer i t s of St andar d Devi at i on
I t i s based on al l t he i t ems of
t he di st r i but i on.
i t i s a meanabl e t o al gebr ai c
t r eat ment si nc e ac t ual + or
si gns devi at i ons ar e t ak en i nt o
c onsi der at i on.
I t i s l east af f ec t ed by
f l uc t uat i ons of sampl i ng
Mer i t s of St andar d Devi at i on
I t f ac i l i t at es t he c al c ul at i on of
c ombi ned st andar d Devi at i on and
Coef f i c i ent of Var i at i on ,w hi c h i s
used t o c ompar e t he var i abi l i t y of
t w o or mor e di st r i but i ons
I t f ac i l i t at es t he ot her st at i st i c al
c al c ul at i ons l i k e sk ew ness
,c or r el at i on.
i t pr ovi des a uni t of measur ement f or
t he nor mal di st r i but i on.
Limitations of Standard Deviation
I t c ant be used f or c ompar i ng
t he var i abi l i t y of t w o or mor e
ser i es of obser vat i ons gi ven i n
di f f er ent uni t s. A c oef f i c i ent of
St andar d devi at i on i s t o be
c al c ul at ed f or t hi s pur pose.
I t i s di f f i c ul t t o c omput e and
c ompar ed
Limitations of Standard Deviation
I t i s ver y muc h af f ec t ed by
t he ex t r eme val ues.
The st andar d devi at i on c an
not be c omput ed f or a
di st r i but i on w i t h open-end
c l asses.

Var i anc e
Var i anc e i s t he ar i t hmet i c mean of
t he squar es of devi at i ons of al l t he
i t ems of t he di st r i but i ons f r om
ar i t hmet i c mean .I n ot her w or ds,
var i anc e i s t he squar e of t he
St andar d devi at i on=

Var i anc e=
2
iance var =
I nt er pr et at i on of Var i anc e
Smal l er t he var i anc e
,gr eat er t he uni f or mi t y i n
popul at i on.
Lar ger t he var i anc e ,gr eat er
t he var i abi l i t y
The Coefficient of Variation
The c oef f i c i ent of var i at i on i s a
measur e of r el at i ve var i abi l i t y
I t i s used t o measur e t he
c hanges t hat have t ak en
pl ac e i n a popul at i on over
t i me
To c ompar e t he var i abi l i t y of
t w o popul at i ons t hat ar e
ex pr essed i n di f f er ent uni t s of
measur ement
I t i s ex pr essed as a
per c ent age

For mul a:
Wher e:

X = mean
= st andar d devi at i on

100
X
CV

1. Dispersion measures
(a) The scatter ness of a set of
observations
(b) The concentration of set of
observations
( c) The Peaked ness of distribution
(d) None

1. Dispersion measures
(a) The scatter ness of a set of
observations
(b) The concentration of set of
observations
( c) The Peaked ness of distribution
(d) None

2. Which one is an absolute measure of
Dispersion?
(a) Range
(b) Mean Deviation
( c) Quartile Deviation
( d) all these measures
2. Which one is an absolute measure of
Dispersion?
(a) Range
(b) Mean Deviation
( c) Quartile Deviation
( d) all these measures
3.Which measures of Dispersion is not
affected by the presence of extreme
observations
(a) deviation
(b) Quartile Deviation
(C) Mean Deviation
(d) Range
observations
(a) deviation
(C) Mean Deviation
(d) Range
observations
(a) deviation
(C) Mean Deviation
(d) Range

4.which measures of Dispersion is
based on all the items of observations
(a) Mean Deviation
(b) Standard Deviation
(C) Quartile Deviation
(d) a and b but not c

4.which measures of Dispersion is
based on all the items of observations
(a) Mean Deviation
(b) Standard Deviation
(C) Quartile Deviation
(d) a and b but not c
5.Standard Deviation is
(a) absolute measure
(b) relative measure
(c) both
(d) none
(a) absolute measure
(b) relative measure
(c) both
(d) none
6.Coefficient of standard deviation is
(a) SD/mean
(b) SD/Median
(c) SD/Mode
(d) Mean/SD

6.Coefficient of standard deviation is
(a) SD/mean
(b) SD/Median
(c) SD/Mode
(d) Mean/SD

7.Coefficient of Quartile Deviation is
calculated by formula
(a)(Q3-Q1)/4
(b) (Q3-Q1)/2
(c) (Q3-Q1)/(Q3+Q1)
(d) (Q3+Q1)/(Q3-Q1

7.Coefficient of Quartile Deviation is
calculated by formula
(a)(Q3-Q1)/4
(b) (Q3-Q1)/2
(c) (Q3-Q1)/(Q3+Q1)
(d) (Q3+Q1)/(Q3-Q1

8.The standard Deviation of
5,8,5,5,5,8and 8 is
(a) 4
(b) 6
(C) 3
(d) 0
8.The standard Deviation of
5,8,5,5,5,8and 8 is
(a) 4
(b) 6
(C) 3
(d) 0
9.If all the observations are increased
by 10,then
(a) SD would be increased by 10
(b)Mean deviation would be increased
by 10
( c) Quartile Deviation would be
increased by 10
(d) all these remain unchanged
9.If all the observations are increased
by 10,then
(a) SD would be increased by 10
(b)Mean deviation would be increased
by 10
( c) Quartile Deviation would be
increased by 10
(d) all these remain unchanged
10.For any two numbers SD is always
(a) Twice the range
(b) Half of the range
Square of range
(d) none of these
10.For any two numbers SD is always
(a) Twice the range
(b) Half of the range
Square of range
(d) none of these
11Mean deviation is minimum when
deviations are taken about
(a) Arithmetic Mean
(b) Geometric Mean
Harmonic Mean
(d) Median

11.Mean deviation is minimum when
deviations are taken about
(a) Arithmetic Mean
(b) Geometric Mean
Harmonic Mean
(d) Median

12.Root mean square deviation is
(a) Standard Deviation
both
(d) none

12.Root mean square deviation from
mean is
(a) Standard Deviation
both
(d) none

(a) Smaller than mean deviation about
mean
(b) Smaller than mean deviation about
median
Larger than mean deviation about mean
(d) none of these

(a) Smaller than mean deviation about
mean
(b) Smaller than mean deviation about
median
Larger than mean deviation about mean
(d) none of these

14.Is least affected by sampling
fluctuations
a) Standard Deviation
both
(d) none

14.Is least affected by sampling
fluctuations
a) Standard Deviation
both
(d) none

15. Coefficient of variation of two series is 60%
and 80% respectively. Their standard deviations
are 20 and 16 respectively, what is their A.M
A) 15 and 20
B) 33.3 and 20
C) 33.3 and 15
D) 12 and 12.8

15. Coefficient of variation of two series is 60%
and 80% respectively. Their standard deviations
are 20 and 16 respectively, what is their A.M
A) 15 and 20
B) 33.3 and 20
C) 33.3 and 15
D) 12 and 12.8

16. For the numbers 1, 2, 3, 4, 5, 6, 7 standard
deviation is:
A) 3
B) 4
C) 2
D) None of these
16. For the numbers 1, 2, 3, 4, 5, 6, 7 standard
deviation is:
A) 3
B) 4
C) 2
D) None of these
THE END
Measures of Dispersion

Measures of Dispersion: Quantitative Aptitude & Business Statistics

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Measures of Dispersion: Quantitative Aptitude & Business Statistics

Caricato da

Copyright:

Formati disponibili

Measures of Dispersion

Quantitative Aptitude & Business Statistics

Quantitative Aptitude & Business Statistics:

Quantitative Aptitude & Business Statistics:

Quantitative Aptitude & Business Statistics:

Quantitative Aptitude & Business Statistics:

Quantitative Aptitude & Business Statistics:

Quantitative Aptitude & Business Statistics:

Potrebbero piacerti anche