THE MEASURES OF DISPERSION

Nahida Sultana
Lecturer
Business Administration
BU
WHY STUDY DISPERSION ???
An average, such as the mean
or the median only locates the
center of the data.
An average does not tell us

anything about the spread of the

data.
WHAT IS DISPERSION ?
Dispersion ( also known as
Scatter , spread or variation )
measures the items vary from
some central value.
It measures the degree of

variation.
Ages: 18 20 22 20 23 22 24 25 26 25 23 24
Mean: 272/12 =22.6=23

Ages Dispersion (variation)

18 5
20 3
22 1
20 3
23 0
22 1
24 1
25 2
23 0
24 1
SIGNIFICANCE OF MEASURING
DISPERSION :
To determine the reliability of an
average.
 To facilitate comparison.

 To facilitate control.

To facilitate the use of other

statistical measures.
PROPERTIES OF GOOD MEASURES OF
DISPERSION :
 Simple to understand and easy to
calculate.
 Rigidly defined.
 Based on all items.
 A meanable to algebric
treatment.
 Sampling stability.
 Not unduly affected by extreme
items.
 Absolute
measure of
dispersion

Based on
Selected Based on
Items all Items

Inter Mean Standard

Range Quartile
Range deviation deviation
 Relative measure of
Dispersion

Based on
Based on all Items
Selective Items

Coefficient Coefficient Coefficient Coefficient Coefficient

of range of QD of MD of SD of Variation
METHODS OF STUDYING VARIATION :
 The Range
 The Inter Quartile Range and the Quartile Deviation

 Mean Deviation

 Standard Deviation
THE RANGE
 The simplest measure of dispersion is the range.
 Foe ungrouped data , the range is the difference
between the highest and the lowest value in a set of
data.

RANGE = HIGHEST VALUE – LOWEST VALUE

RANGE

Rice Wheat Cereals Pulses Total

81.12 87.86 30.324 10.90 200.73

Range = 87.86 – 10.90

= 76.96

Co-efficient of range=
(87.86 – 10.90)/ (87.86 +10.90)
=76.96/98.76
=0.78
COEFFICIENT OF RANGE :
 Coefficient of Range is calculated as ,

Coefficient of Range= L-S

L+S
COEFFICIENT OF RANGE

Rice Wheat Cereals Pulses Total

0.613 0.846 0.343 0.392 0.609

MERITS OF RANGE

Simple to understand and the

easiest to compute.
Takes less time to calculate.
It does not require any special

knowledge.
LIMITATIONS OF RANGE :
 Range is not based on each and every item
of distribution.
 It takes in to account only the most extreme
values.
 Range can not computed be in case of open
end distribution.
 It does not indicate the direction of
variability.
INTERQUARTILE RANGE
 It is the second measure of dispersion, no doubt
improved version over the range.
 It is based on the quartiles so while calculating this
may require upper quartile (Q3) and lower quartile
(Q1) .
 It is the deference between two quartiles

Interquartile range = Q3 – Q1
18 20 22 20 23 22 24 25 26 25 23 24

Data arranged in Ascending order:

18 18
20 38
20 58 Q1
22 80
22 102
23 125
IQR
23 148
24 172
24 196 Q3
25 221
25 246
26 272

Q1 = N/4 th observation = 272/4 = 68th observation

Q3 = 3N/4 th observation = (3*272)/4 = 204th observation
IQR = (Q3- Q1)= (204 – 68 ) = 136
Co-efficient of QD: 136/272 =0.5
INTER QUARTILE RANGE :

Rice Wheat Cereals Pulses Total

Q1 = 40.82 Q1 = 23.89 Q1 = 26.08 Q1 = 11.39 Q1 = 99.83

Q3 = 84.51 Q3 = 69.35 Q3 = 33.52 Q3 = 13.84 Q3 = 199.43

43.71 45.453 7.442 2.448 99.6

COEFFICIENT OF QUARTILE DEVIATION

 Quartile Deviation = Q3 – Q1
Q3 + Q1

Rice Wheat Cereals Pulses Total

Q1 = 40.82 Q1 = 23.89 Q1 = 26.08 Q1 = 11.39 Q1 = 99.83

Q3 = 84.51 Q3 = 69.35 Q3 = 33.52 Q3 = 13.84 Q3 = 199.43

0.348 0.487 0.124 0.0969 0.332

MERITS OF QUARTILE DEVIATION:
 It can be easily calculated and simply understood.
 It does not involve much mathematical difficulties.

 As it takes middle 50% terms hence it is a measure

better than Range and Percentile Range.
 It is not affected by extreme terms as 25% of upper
and 25% of lower terms are left out
LIMITATIONS OF QUARTILE DEVIATION
 It is not suited to algebraic treatment.
 It is very much affected by sampling
fluctuations.
 The method of Dispersion is not based
on all the items of the series.
 It ignores the 50% of the distribution.
 MEAN DEVIATION :
 The mean deviation takes into consideration all of
the values.
 Mean Deviation : The arithmetic mean of the
absolute values of the deviations from the
arithmetic mean
x= value of each observation

x= the arithmetic mean of the

values

n = the no of observations.
MEAN DEVIATION :

Rice Wheat Cereals Pulses Total

21.76 23.14 4.44 1.71 50.26

MERITS OF MEAN DEVIATION
 It is easy to understand
 It is based on all the items of the
series.
 It is less affected by extreme
values.
 It is useful in small samples when
no detailed analysis is required
LIMITATIONS OF MEAN DEVIATION
 It lack properties such that ( + ) and
( - ) signs which are not taken in to
consideration.
 It is not suitable for mathematical
treatment.
 It may not give accurate results when the
degree of variability in a series is very high.
STANDARD DEVIATION
 Standard deviation is the most commonly
used measure of dispersion.
 Similar to the mean deviation , the standard
deviation takes in to account the value of
every observation.
 The values of the mean deviation and the
standard should be relatively similar.
 It is a measure of spread of data about the
mean. SD is the square root of sum of
squared deviation from the mean divided by
the number of observations.
STANDARD DEVIATION :

Rice Wheat Cereals Pulses Total

24.43 26.69 5.71 61.48 57.77

PROPERTIES OF STANDARD DEVIATION
 Independent of change of origin.
 Not independent of change of scale.
 Fixed relationship among measures of
Dispersion.
MERITS OF STANDARD DEVIATION :
 It is based on all the items of the distribution.
 It is meanable to algebraic treatment since actual +
or – signs deviations are taken into consideration.
 It is least affected by fluctuations of sampling.

 It provides a unit of measurement for the normal

distribution.
LIMITATIONS OF STANDARD DEVIATION :
 It can’t be used for comparing the
variability of two or more series of
observations given in different units.
 It is difficult to compute and compared.
 It is very much affected by the extreme
values.
 The standard deviation can not be
computed for a distribution with open end
classes.
THANK
YOU