SDJ INTERNATIONAL COLLEGE, Vesu

Measures of Dispersion

Meaning : "Dispersion is a measure of variation of the items."

In other words "The extent to which a numerical data is likely to vary about an average value is known as dispersion."

Characteristics of an Ideal measures of Dispersion:

 It should be rigidly defined.

 It should not be affected much by sampling fluctuations.
 Its calculation is easy and easy to understand.
 It should be capable of further mathematical treatment.
 It should be based on all observation.

Objectives of Dispersion:

 To find out the average of different values in a distribution from the mean.
 To find out the range of different values in the series.
 To make a comparative study of the variability of the two series.
 To serve as a base for further statistical analysis.

Measures of Dispersion:

Measure of
Dispersion

Absolute Relative
Measures Measures

Absolute Measures:

The measures of dispersion which can be expressed in terms of the original units of a series. Such as rupees, meters,
kilograms etc. are termed as absolute measures.

This measures may be erroneous(wrong or incorrect) for comparison if the series are originally in different units.

Relative Measures:

The Measures of dispersion which can be expressed as a percentage or ratio of the average are called relative
measures.

For comparing the variability of two or more series it is the relative measures that has to be taken into account.

Prepared By : Prof. Rajeev Yadav

 SDJ INTERNATIONAL COLLEGE, Vesu

Methods of Measuring Dispersion

Methods of
Measuring
Dispersion

Methods of
Methods of
Averaging
Limits
Deviations

Quartile Mean Standard

Range
Deviation Deviation Deviation

Range:

It is the simplest measure of Dispersion. It is the difference between the values of the extreme items of a series. Range
is defined as the difference between the highest and the lowest values taken by the variables.

Range (R) = Highest Value (H) - Lowest Value (L)

Coefficient of Range:

To compare the variability of two or more distributions given in the different units of measurement, the relative
measure of range is called coefficient of range.
𝐻−𝐿
𝐶𝑜𝑒𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑅𝑎𝑛𝑔𝑒 = 𝐻+𝐿

[Note: Range deals with the value of observation and not with the number of occurrence in the data set. So
frequency is not taken into consideration]

Merits of Range:

 It is easily calculated and readily understood.

 It gives a broad picture of the data and gives the limits within all the items fall.
 It is rigidly defined.

Demerits of Range:

 It is affected greatly by fluctuation of sampling.

 It is not based on all the observation of the series.
 It is not suitable for mathematical treatment.
 It is not treated as reliable measure of variability since it is based on two extreme values.

Uses of Range:

 Quality Control: The control charts are prepared on the basis of length for controlling the quality of the
products.
 Measurement of Fluctuations: It is a useful measure to study variations of the data in our day to day life.
 Whether Forecasting: It is used for weather forecasting since the people are interested to know the limits of
temperature.

Prepared By : Prof. Rajeev Yadav

 SDJ INTERNATIONAL COLLEGE, Vesu
Quartile Deviation:

Quartile deviation is the mid-point of the range between the two quartiles. It is half of Inter-Quartile Range. It is also
called semi-inter quartile range.

𝑄3 − 𝑄1
𝑄𝑢𝑎𝑟𝑡𝑖𝑙𝑒 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑄𝐷 =
2

Coefficient of Quartile Deviation:

For comparing study between two or more variable, its relative measure called coefficient of quartile deviation is
computed.
𝑄3 − 𝑄1
𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑄𝐷 =
𝑄3 + 𝑄1

Merits of Quartile Deviation:

 It is simple to calculate and easy to understand.

 It is not affected by extreme values, as it gives spread of the control items only.
 It is useful in dealing with open-end distribution.

Demerits of Quartile Deviation:

 It is not a measure of dispersion since it does not study the dispersal of values from a measure of central
tendency.
 It is affected considerably by fluctuations of sampling.
 It is not suitable for further mathematical treatment.
 It is not based on all the observations of the data as it ignores the first 25% and the last 25% of the items.

Mean Deviation:

Mean Deviation is the measure of dispersion which is based upon all the items in a variable.

In other words, "Average Deviation is the average amount of scatter of items in a distribution from either mean
or the median, ignoring signs of the deviation."

Data Type From Mean From Median From Mode

Raw Data 𝑥𝑖 − 𝑥 𝑥𝑖 − 𝑀 𝑥𝑖 − 𝑀0
𝑀𝐷 = 𝑀𝐷 = 𝑀𝐷 =
𝑛 𝑛 𝑛

Grouped data 𝑓 𝑥𝑖 − 𝑥 𝑓 𝑥𝑖 − 𝑀 𝑓 𝑥𝑖 − 𝑀0
𝑀𝐷 = 𝑀𝐷 = 𝑀𝐷 =
𝑛 𝑛 𝑛

Continuous Data 𝑓 𝑥𝑖 − 𝑥 𝑓 𝑥𝑖 − 𝑀 𝑓 𝑥𝑖 − 𝑀0
𝑀𝐷 = 𝑀𝐷 = 𝑀𝐷 =
𝑛 𝑛 𝑛

Prepared By : Prof. Rajeev Yadav

 SDJ INTERNATIONAL COLLEGE, Vesu

Coefficient of Mean Deviation:

The purpose of comparing variations or dispersal between two or more different sets of data (In different units or same
units), a relative measure is computed using the following formulae:

Coefficient of Mean Deviation:

MD from Mean MD from Median MD from Mode

𝑀𝐷 𝑀𝐷 𝑀𝐷
𝑥 𝑀 𝑀0

Merits of Mean Deviation:

 It is simple to calculate and easy to understand.

 It takes into account all the items of a series. Hence, it is affected by every value in the distribution.
 Mean Deviation can be calculated from any average, viz mean, median or mode. But the use of mean and
median are very popular.
 It is not affected by extreme values.
 Mean deviation is rigidly defined and it is rigid and calculated values.

Demerits of Mean Deviation:

 Mean deviation ignores the algebraic signs of the deviations, and as such, it is not capable of further algebraic
treatment.
 It is not a very accurate measure of dispersion particularly when it is calculated from the mode because mode
can be unrepresentative.

Uses of Mean Deviation:

 It is used for forecasting of business cycles, and also for small samples studies.

Standard Deviation:

Standard deviation is the square root of the arithmetic average of the squares of the deviations measured from the
mean.

Standard Deviation(SD) is represented by a Greek Letter Sigma ( σ ).

𝑥𝑖 − 𝑥 2
𝑆𝐷 =
𝑛

Coefficient of Standard Deviation:

Standard Deviation is an absolute measure. Where comparison of variability in two or more series is required to be
made, relative measure of standard deviation is compared. It is called coefficient of standard deviation.

𝜎
𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑜𝑓 𝑆𝐷 =
𝑥

Prepared By : Prof. Rajeev Yadav

 SDJ INTERNATIONAL COLLEGE, Vesu
Data Type Formula Points to be considered
Raw Data 𝑥𝑖 − 𝑥 2 When 𝑥 is integer or
𝑆𝐷 = whole number.
𝑛

𝑥2 𝑥2 𝑥 2
When 𝑥 is non integer or
𝑆𝐷 = − 𝑥 2 or 𝑆𝐷 = − irrational number.
𝑛 𝑛 𝑛
𝑑 2 When 𝑥 is non integer or
𝑑2
𝑆𝐷 = − 𝑛 irrational number.
𝑛
Grouped data 𝑓 𝑥𝑖 − 𝑥 2 When 𝑥 is integer or
𝑆𝐷 = whole number.
𝑛

𝑓𝑥 2 𝑓𝑥 2 𝑓𝑥 2
When 𝑥 is non integer or
𝑆𝐷 = − 𝑥2 or 𝑆𝐷 = − irrational number.
𝑛 𝑛 𝑛
𝑓𝑑 2 When 𝑥 is non integer or
𝑓𝑑 2
𝑆𝐷 = − 𝑛 irrational number.
𝑛
Continuous Data 𝑓 𝑥𝑖 − 𝑥 2 When 𝑥 is integer or
𝑆𝐷 = whole number.
𝑛

𝑓𝑥 2 𝑓𝑥 2 𝑓𝑥 2
When 𝑥 is non integer or
𝑆𝐷 = − 𝑥 2 or 𝑆𝐷 = − irrational number.
𝑛 𝑛 𝑛
𝑓𝑑 2 When 𝑥 is non integer or
𝑓𝑑 2
𝑆𝐷 = − 𝑛 irrational number.
𝑛
Combined Standard When we have to find
𝑛 1 𝜎12 +𝑛 2 𝜎22 + 𝑛 3 𝜎32 +⋯𝑛 𝑘 𝜎𝑘2
Deviation 𝜎𝑐 = the combined standard
𝑛 1 +𝑛 2 +𝑛 3 +⋯ 𝑛 𝑘 deviation.

Merits of Standard Deviation:

Based on all items: Standard deviation is based on all items of a variable. It makes use of all the information
found in the set of data.
Well-defined and definite: Standard deviation is rigidly defined and definite measure of dispersion. It is
obtained in any condition.
Algebraic properties: It is the easiest measure to handle algebraically. It is the result of correct mathematical
process.
Least affected by sample variations: It is not affected by the fluctuations of sampling and is of immense
importance for sampling theory.
Useful and significant: It is a significant measure of dispersion for making comparisons between variability
or consistency of two sets of observations and to test the significance of other statistical measures like
correlation and regression analysis.
Convenient for interpretation: It provides unit of measurement for the study of normal distribution since it
is convenient for interpretation.

Demerits of Standard Deviation:

Difficult calculation: Standard deviation is comparatively difficult to calculate and not easy to understand.
Calculation of squares of deviations and then obtaining their mean and square root are not easy arithmetical
processes.
Affected by extreme values: The extreme values are given more weight in its calculation, since values are
squared up.
Prepared By : Prof. Rajeev Yadav