Measures of Dispersion

Dispersion: Dispersion measures the extent to which individual items differs. It indicates
lack of uniformity in the size of items. It is also called variation or variability. The
measures of dispersion are called average of second order.
Following are the measures of dispersion:

A. Absolute Measures:
1. Range
2. Mean Deviation(MD)
3. Standard Deviation(SD)
4. Quartile Deviation(QD)
B. Relative Measures:
1. Coefficient of range
2. Coefficient of mean deviation
3. Coefficient of Standard Deviation
4. Coefficient of quartile deviation
A. Absolute Measure: When we measure the dispersion and express it in terms of
original data, it is called an absolute measure.
B. Relative Measure: Absolute measure cannot be used for comparison purposes if
expressed in different units. In order to compare two series, if absolute measure of
dispersion of each series is expressed as a ratio or percentage of the average then
itg is called relative measure of dispersion.

1. Range: Range is the difference between highest and lowest observation in a set of
data.
Calculation of Range:
Ungrouped Data: Let, x1 , x 2 ,......, x n are n observations. x H is the highest value and x L
is the lowest value.
So, Range = x H − x L
Grouped Data: Let, x1 , x 2 ,......, x n are the middle value of the class intervals and their
corresponding frequencies are f1 , f 2 ,......, f n .Then-
Range=Upper limit of last class interval –lower limit of first class interval.
Coefficient of range:
For ungrouped data:
Highest value − lowest value
Coefficient of range= (if percentage wanted then multiply
Highest value + lowest value
by 100)
x − xL
= H
xH + xL

By Samsad Jahan
For grouped data:
Coefficient of range
Upper limit of last class interval − Lower limit of first class interval
=
Upper limit of last class interval + Lower limit of first interval
x − xL
= U
xU + x L

Example 1: Find the range and coefficient of range in the series:

80, 90, 60, 63, 68, 61, 67, 65, 100, 75, 89, 84, 86

Example 2:From the following data find the range and coefficient of range

Profit in Lac Taka Number of companies

25-35 4
35-45 7
45-55 15
55-65 14
65-75 8
75-85 2

Uses: The range is most widely used in statistical process control (SPC) applications
because it is very easy to calculate and understand.

2. Mean Deviation: Mean deviation is the arithmetic mean of the absolute deviation of
observations from any measure of central tendency (mean, median, and mode).

Calculation of mean deviation:

Ungrouped Data: Let, x1 , x 2 ,......, x n are n observations with mean x , then mean
deviation about mean is given by-
n
 xi − x
MD( x )= i =1
n
Let, x1 , x 2 ,......, x n are n observations with median M e , then mean deviation about
median is given by-
n
 xi − M e
MD( M e )= i =1
n
Let, x1 , x 2 ,......, x n are n observations with mode M 0 , then mean deviation about Mode
is given by-
n
 xi − M 0
MD ( M 0 ) = i =1
n

By Samsad Jahan
Grouped Data: Let, x1 , x 2 ,......, x n are the middle value of the class intervals and their
corresponding frequencies are f1 , f 2 ,......, f n and their mean is x .Then-
n
 f i ( xi − x )
Mean deviation about mean, MD( x )= i =1
N
Similarly,
n
 f i ( xi − M e )
Mean deviation about median, MD( M e )= i =1
N
n
 f i ( xi − M 0 )
Mean deviation about mode, MD( M 0 )= i =1
N
Coefficient of mean deviation:
MD( x )
Coefficient of mean deviation about mean, CMD( x ) =  100
x
MD( M e )
Coefficient of mean deviation about median, CMD( M e ) =  100
Me
MD( M 0 )
Coefficient of mean deviation about mode, CMD( M 0 ) =  100
M0

Advantages:
1. It can be easily understood
2. It utilizes all the observations
Disadvantages:
1.It is not a satisfactory measure if the series is not symmetrical.
2. It is suitable for algebraic treatment.
Ex: - The number of patients seen in the emergency room at Square Hospital for a sample
of 5 days last year was-
103, 97, 101, 106, 103
Determine the mean deviation about mean, median and mode. Also find the relative
measures in percentage.
3.Standard deviation: The arithmetic mean of the square deviations of observations
from their arithmetic mean is known as variance. The positive square root of variance is
known as standard deviation.
Symbols related to Standard deviation
Mean Standard Deviation Denominator
Population   N
Sample x s n − 1 ,Ungrouped data
N − 1 ,Grouped data

By Samsad Jahan
Calculation of standard deviation:
For ungrouped sample data:

Sample mean, x =
 xi
n
Let x1 , x2 ,........, xn be n observations with mean x , then standard deviation is given by-
n
 ( xi − x )
2

Sample Standard Deviation, s = i =1

n −1
n
( xi2 − nx 2 )
Or, s = i =1

n −1
n
(  x − nx ) 2
i
2

Sample variance, s 2 = i =1

n −1
For grouped sample data:
n

fx i i
Sample Mean, x = i =1

N
Let, x1 , x 2 ,......, x n are the middle value of the class intervals and their corresponding
frequencies are f1 , f 2 ,......, f n and their mean is x .Then standard deviation is given by-
n

 f (x i i − x)2
Sample Standard Deviation , s = i =1

N −1
n
( f i xi2 − Nx 2 )
Or, s = i =1

N −1
n

fx i
2
i − N x2
Sample variance, s 2 = i =1

N −1

Ungrouped Population Data:

N
 xi
population mean,  = i =1

N
N
 ( xi −  )
2

Population Standard deviation:  = i =1

By Samsad Jahan
 N
 ( xi −  )
2

Population variance,  2 = i =1

N
Grouped Population Data:
n

fx i i
Population mean:  = i =1

N
n

fx i
2
i − N 2
Population Variance,  2 = i =1

Population standard deviation,  =

fx i
2
i − N 2
N

Advantages: i) It utilizes all observations.

ii) It is less erratic.
iii) Suitable foe algebraic manipulation.
iv) It is used for higher statistical operations.
v) Less affected by fluctuation of sampling
Disadvantages:
i) Its calculation demands greater time and labour.
ii) It gives greater weights to extreme items.

Uses of Standard Deviation:

As a measure of dispersion standard deviation is the most important. An important
application of SD is to judge how much representative the AM is as a measure of central
tendency. Thus if two or more comparable distribution have the same mean, Then the
mean is the most representative in the distribution with smallest SD. It is useful for
locating the place of an individual item in the distribution. It is a key stone in sampling
and correlation and also used in interpretation of curves.

Coefficient of standard deviation or coefficient of variation:

The ratio of the standard deviation to the arithmetic mean expressed as a percentage.
It is a very useful measure when-
i) The data are in different units (such as dollar and days absent)
ii) The data are in the same unit, but the mean are far apart.( The income
of the top executives and the income of the unskilled employees).
s
Now- Coefficient of variation, C.V=  100 (for sample)
x

C.V=  100 (for population)

Where,  =population standard deviation

By Samsad Jahan
 N
 xi
Population mean,  = i =1

Quartile Deviation: Quartile deviation or semi-interquartile range is obtained by

dividing the difference between the upper and lower quartile by 2.
Upperquartile − lowerquartile
i.e. Quartile Deviation or Q.D=
2
Q3 − Q1
=
2
Where, For ungrouped data-
n +1
Q1 = th observation in a series arranged in ascending order
4
3(n + 1)
Q3 = th observation in a series arranged in ascending order.
4
For grouped Data-
N
First quartile group= size of th item
4
N − p.c. f
Q1 = L1 + 4  c , Where L1 =lower limit of the first quartile group
f1
N =Total number of observation
p. c. f = Cumulative frequency of the group just
preceding the 1st quartile group.
f1 =Frequency of the first quartile group
c = Class interval
3N
Third quartile group= size of th item
4
3N − p.c. f
Q3 = L3 + 4  c , Where L3 =lower limit of the Third quartile group
f3
N =Total number of observation
p. c. f = Cumulative frequency of the group just
preceding the 3rd quartile group.
f 3 =Frequency of the third quartile group
c = Class interval
Second quartile, Q2 =Median
Q − Q1
Coefficient of Quartile Deviation= 3
Q3 + Q1
Samsad Jahan
Assistant Professor of Statistics
Department of Arts and Sciences
Ahsanullah University of Science and Technology

By Samsad Jahan
By Samsad Jahan