BECM/Math-2207 (MHU) 1

Lecture: 3-4: Measures of Dispersion

Definition: If we carefully observed a series of data, we find that the observation very among themselves and
also form the average. While studying a dispersion it is equally important to know how the variables are
clustered around or scattered a way from the point of central tendency. If two distribution centers round the
same point viz the mean, yet differ in variously mean such variation is called dispersion or spread or scatter or
variability.
Dispersion is absolute measured:(i) Range (ii) Quartile deviation (iii) Mean deviation (iv)Standard deviation.
Range: The range is the difference between two extreme observations of the distribution. If A and B are the
greatest and smallest observations respectively in a distribution then its range is A-B.
Quartile deviation: Quartile deviation or semi-inter quartile range Q is given by 𝑄 = (𝑄 − 𝑄 ).

Where Q1 and Q3 are the first and third quartiles of the distribution respectively.

Mean deviation: For the frequency distribution 𝑥 |𝑓 , (𝑖 = 1, 2, 3, ⋯ ⋯ , 𝑛), the mean deviation from the

average A (usually mean, median or mode) is given by Mean deviation= ∑ 𝑓 |𝑥 − 𝐴|, where ∑ 𝑓 = 𝑁.

Thus, Mean deviation from mean= ∑ 𝑓 |𝑥 − 𝑥̄ |

Mean deviation from median= ∑ 𝑓 |𝑥 − 𝑚𝑒𝑑𝑖𝑎𝑛|

Mean deviation from mode= ∑ 𝑓 |𝑥 − 𝑚𝑜𝑑𝑒|

Standard deviation: Standard deviation, usually denoted by the Greek letter 𝜎 (sigma), is the square root of
the arithmetic mean of the squares of the deviations of the given values from their arithmetic mean. For the

frequency distribution 𝑥 |𝑓 , (𝑖 = 1, 2, 3, ⋯ ⋯ , 𝑛), standard deviation, 𝜎 = ∑ 𝑓 (𝑥 − 𝑥̄ ) .

where 𝑥̄ is the arithmetic mean of the distribution and ∑ 𝑓 = 𝑁.

Variance: The square of standard deviation is called the variance and is given by 𝜎 = ∑ (𝑥 − 𝑥̄ ) .

For the frequency distribution 𝑥 |𝑓 , (𝑖 = 1,2,3, ⋯ ⋯ , 𝑛), variance, 𝜎 = ∑ 𝑓 (𝑥 − 𝑥̄ ) .

where 𝑥̄ is the arithmetic mean of the distribution and ∑ 𝑓 = 𝑁.

Root mean square deviation: Let, 𝑥 , (𝑖 = 1, 2, 3, ⋯ ⋯ , 𝑛) be n observations and A be any arbitrary

number, Root mean square deviation is denoted by ‘S’ is given by 𝑆 = ∑ (𝑥 − 𝐴) .

For the frequency distribution 𝑥 |𝑓 , (𝑖 = 1, 2, 3, ⋯ ⋯ , 𝑛), 𝑆 = ∑ 𝑓 (𝑥 − 𝐴) .where ∑ 𝑓 = 𝑁.

Mean square deviation: The square of the root mean square deviation is called the mean square deviation
and denoted by ‘S2’ is given by 𝑆 = ∑ (𝑥 − 𝐴) .
For the frequency distribution 𝑥 |𝑓 , (𝑖 = 1, 2, 3, ⋯ ⋯ , 𝑛),
variance, 𝑆 = ∑ 𝑓 (𝑥 − 𝐴) , where ∑ 𝑓 = 𝑁.
BECM/Math-2207 (MHU) 2

Coefficient of dispersion: The coefficient of dispersion (C.D) based on different measures of dispersion are
as follows: (i) Based upon range, C.D = , A and B are the greatest and the smallest terms in the series.

(ii) Based upon mean deviation: C.D = ,

ℎ ℎ

(iii) Based upon quartile deviation: C.D= ,

(iv) Based upon standard deviation: C.D= .

̄
Coefficient of variation: 100 times the coefficient of dispersion based upon standard deviation is called
coefficient of variation, i.e, 𝐶. 𝑉 = 100 × ̄ .

Question-1: Which measures is the best measures of dispersion?

Answer: Among the four absolute measures of dispersion, standard deviation is the best measures of dispersion,
because it posses almost all the requisites of a good measure of dispersion.
(i) It is rigidly defined.
(ii) It is based on all the observations even if one of the observations is changed standard deviation changed.
However, range, Quartile deviation does not possess this property.
(iii) It is easy to understand and easy to calculate.
(iv) It is amenable to algebraically treatment.
(v) It is affected least by fluctuation of sampling.

Question-2: Determine the relation between standard deviation and root mean square deviation i.e, 𝝈 and𝑺.
Solution: By definition we have 𝑆 = ∑ 𝑓 (𝑥 − 𝐴) ,

= ∑ 𝑓 (𝑥 − 𝑥̄ + 𝑥̄ − 𝐴) ,

= ∑ 𝑓 [(𝑥 − 𝑥̄ ) + (𝑥̄ − 𝐴) + 2(𝑥 − 𝑥̄ )(𝑥̄ − 𝐴)],

= ∑ 𝑓 (𝑥 − 𝑥̄ ) + (𝑥̄ − 𝐴) ∑ 𝑓 +2 ∑ 𝑓 (𝑥 − 𝑥̄ )(𝑥̄ − 𝐴),

= ∑ 𝑓 (𝑥 − 𝑥̄ ) + (𝑥̄ − 𝐴) + 0, [Since sum of the deviation of A.M=0]

or,𝑆 = 𝜎 + (𝑥̄ − 𝐴)
or,𝑆 = 𝜎 + 𝑑 which is the required relation. where 𝑑 = 𝑥̄ − 𝐴.
Obviously S2 will be least when d=0 i.e, 𝑥̄ = 𝐴.
Hence mean square deviation and consequently root mean square deviation is least when the
deviations are taken from 𝐴 = 𝑥̄ i.e, standard deviation is the least value of root mean square deviation.

Question-3: Prove that the variance of the first n natural number is .

OR, Prove that the standard deviation of the first n natural number is .

Proof: Let 𝑥 , 𝑥 , 𝑥 , ⋯ ⋯ 𝑥 be n observations with corresponding the first n natural number 1, 2, 3, - - -, n.

We know that, Arithmetic mean, 𝑥̄ = ∑ 𝑥

 BECM/Math-2207 (MHU) 3

= (𝑥 + 𝑥 + 𝑥 + ⋯ ⋯ + 𝑥 )

= (1 + 2 + 3 + ⋯ ⋯ + 𝑛)
( )
=

Again, ∑ 𝑥 = (𝑥 + 𝑥 + 𝑥 + ⋯ ⋯ ⋯ + 𝑥 )

= (1 + 2 + 3 + ⋯ ⋯ + 𝑛 )
( )( )
=
( )( )
=

We know, 𝑉(𝑥) = ∑ (𝑥 − 𝑥̄ )

= ∑ 𝑥 − 𝑥̄
( )( )
= −

= −

= (𝑛 + 1)(𝑛 − 1)

∴ 𝑉(𝑥) =

OR, We know, 𝜎 = ∑ (𝑥 − 𝑥̄ )

= ∑ 𝑥 − 𝑥̄

( )( )
= −

= −

= (𝑛 + 1)(𝑛 − 1)

∴𝜎= (Proved).

∑ ∑
H.W. Question-4: Prove that 𝑆 = − = 𝑥 − 𝑥̄ .
BECM/Math-2207 (MHU) 4

Problem-5: Find the mean deviation from the mean and variance of the series: a, a+d, a+2d,- - -,a+2nd
Solution: Given that the series a, a+d, a+2d, - - - - - - , a+2nd
1st Part: We know, Mean deviation from mean= ∑ |𝑥 − 𝑥̄ |
∑
= .................. (i) where 𝛿 = |𝑥 − 𝑥̄ |
( )( )
Here, mean, 𝑥̄ = = = 𝑎 + 𝑛𝑑 where 𝑁 = + 1 = 2𝑛 + 1
.
𝑆 = {2𝑎 + (𝑁 − 1)𝑑}
and 𝛿 = |𝑥 − 𝑥̄ | = |𝑎 − 𝑎 − 𝑛𝑑| = 𝑛𝑑 = {2𝑎 + (2𝑛 + 1 − 1)𝑑}
𝛿 = |𝑥 − 𝑥̄ | = |𝑎 + 𝑑 − 𝑎 − 𝑛𝑑| = (𝑛 − 1)𝑑 = (2𝑛 + 1)(𝑎 + 𝑛𝑑)
𝛿 = |𝑥 − 𝑥̄ | = |𝑎 + 2𝑑 − 𝑎 − 𝑛𝑑| = (𝑛 − 2)𝑑
Similarly, 𝛿 = |𝑥 − 𝑥̄ | = |𝑎 + (𝑛 − 1)𝑑 − 𝑎 − 𝑛𝑑| = 𝑑
𝛿 = |𝑥 − 𝑥̄ | = |𝑎 + 𝑛𝑑 − 𝑎 − 𝑛𝑑| = 0
𝛿 = |𝑥 − 𝑥̄ | = |𝑎 + (𝑛 + 1)𝑑 − 𝑎 − 𝑛𝑑| = 𝑑
𝛿 = |𝑥 − 𝑥̄ | = |𝑎 + 2𝑛𝑑 − 𝑎 − 𝑛𝑑| = 𝑛𝑑
∑ ⋯⋯
Now from (i), we get Mean deviation from mean = =
( ) ( ) ⋯⋯ ⋯⋯
=
( ⋯⋯ ( ) )
=
( ⋯⋯ )
=
( )
= ×
( )
= Ans.
2nd Part: we know, Variance, 𝜎 = ∑ (𝑥 − 𝑥̄ ) . = ∑ 𝛿 = (𝛿 +𝛿 +𝛿 + ⋯⋯+ 𝛿 )
( ) ( ) ⋯⋯ ⋯⋯
=
( ⋯⋯ )
=
( )( )
= ×
( )
= Ans.
 BECM/Math-2207 (MHU) 5

Problem-6: Calculate the mean and standard deviation for the following data:
Size of item 6 7 8 9 10 11 12
Frequency 3 6 9 13 8 5 4

Solution: Compute the following table:

Size of item (x) Frequency (f) d=x-a fd f d2
6 3 -3 -9 27
7 6 -2 -12 24
8 9 -1 -9 9
9 13 0 0 0
10 8 1 8 8
11 5 2 10 20
12 4 3 12 36
𝑓 = 𝑁 = 48. 𝑓𝑑 𝑓𝑑 = 124
= 0.
Assumed mean=9,
∑
Mean= 𝑥̄ = 𝑎 + = 9 + 0 = 9Ans.

∑ ( ̄)
Standard deviation=𝑆. 𝐷 = ∑
.

∑ ∑
= ∑
− ∑

= = 1.6 Ans.

Problem-7: Calculate the mean and standard deviation for the following data:
Age(in year) 20-30 30-40 40-50 50-60 60-70 70-80 80-90
No. of people 3 61 132 153 140 51 2

Solution: From the given data compute the following table:

Age No. of people (f) x d=(x-a)/h d2 fd f d2
20-30 3 25 -3 9 -9 27
30-40 61 35 -2 4 -122 244
40-50 132 45 -1 1 -132 132
50-60 153 55 0 0 0 0
60-70 140 65 1 1 140 140
70-80 51 75 2 4 102 204
80-90 2 85 3 9 6 18

𝑓=𝑁 𝑓𝑑 𝑓𝑑

= 542. = −15 = 765

BECM/Math-2207 (MHU) 6

Assumed mean=55,
∑
Mean= 𝑥̄ = 𝑎 + × ℎ = 55 + × 10 = 54.7232Ans.

∑ ∑
Standard deviation= 𝑆. 𝐷 = ∑
− ∑

= − = 1.188 Ans.

H.W.Problem-8: Find the mean deviation and standard deviation for the following data:
(i) 12, 6, 7, 3, 15, 10, 18, 5 Ans. M.D=4.25, S.D=4.87
(ii) 9, 3, 8, 8, 9, 8, 9, 18. Ans. M.D=2.25, S.D=3.87
H.W.Problem-9: Find the mean deviation and standard deviation of the heights of the 100 students of
RUET:
Hight (inches) 60-62 63-65 66-68 69-71 72-74
No. of students 5 18 42 27 8
Ans: M.D=2.26 inches, S.D= 2.92 inches.

H.W.Problem-10: Find the mean and standard deviation for the wage distribution of the 65 employees:
Wages $250.00-$259.99 260.00-269.99 270.00-279.99 280.00-289.99 290.00-299.99 300.00-309.99310.00-319.99
No. of Employees 8 10 16 14 10 5 2
Ans: A. M=$ 279.77, S.D= $ 15.60

H.W.Problem-11: Compute the standard deviation for the following frequency distribution:
Class interval 0-4 4-8 8-12 12-16
Frequency 4 8 2 1
Ans: S.D= 3.266 .

H.W.Problem-12: The following table shows the marks obtained by 100 candidates in an examination.
Compute the mean, media and standard deviation:
Marks 1-10 11-20 21-30 31-40 41-50 51-60
No. of candidates 3 16 26 31 16 8
Ans: Mean=32, Median=32.61, S.D= 12.36 .
 BECM/Math-2207 (MHU) 7

Measures of Location
Quantitles: Quantitles are those values in a series which divide the total frequency into number of equal parts
when the data is arranged in order of magnitude.
Some important quantitles are : (i) Quartiles (ii) deciles (iii) percentiles.

Quartiles: Quartiles are those values of series which divide the total frequency into four equal parts.
×

For a frequency distribution the quartiles are given by 𝑄 = 𝐿 + × 𝑐, i=1, 2, 3, ......

×
Where 𝐿 =Lower limit of the i-th quartile class.(i-th quartile class is that class which contains th

observations).
𝐹 = Cumulative frequency of the pre i-th quartile class.
𝑓 = frequency of i-th quartile class.
c = length of class of the i-th quartile class.
N = Total number of observation.
Deciles: Deciles are those values of series which divide the total frequency into ten equal parts.
×

For a frequency distribution the deciles are given by 𝐷 = 𝐿 + × 𝑐, i=1, 2, 3, ......

×
Where 𝐿 =Lower limit of the i-th deciles class.(i-th deciles class is that class which contains th

observations).
𝐹 = Cumulative frequency of the pre i-th deciles class.
𝑓 = frequency of i-th deciles class.
c = length of class of the i-th deciles class.
N = Total number of observation.
Percentiles: Percentiles are those values of series which divide the total frequency into hundred equal parts.
×
For a frequency distribution the percentiles are given by 𝐷 = 𝐿 + × 𝑐, i=1, 2, 3, ......

Where 𝐿 =Lower limit of the i-th percentiles class.

×
(i-th percentiles class is that class which contains th observations).

𝐹 = Cumulative frequency of the pre i-th percentiles class.

𝑓 = frequency of i-th percentiles class.
c = length of class of the i-th percentiles class.
N = Total number of observation.
BECM/Math-2207 (MHU) 8

Problem-1: Calculate the quartile coefficient of skewness for the following data:
x 3-7 8-12 13-17 18-22 23-27 28-32 33-37 38-42
f 2 108 580 175 80 32 18 5

Solution: From the given data compute the following table:

x f F
3-7 2 2
8-12 108 110
13-17 580 690
18-22 175 865
23-27 80 945
28-32 32 977
33-37 18 995
38-42 5 1000
×
×
First quartiles, 𝑄 = 𝐿 + × 𝑐 [here = = 250occurred in 13-17]
= 13 + ×4 [ L1= 13, F1=110, f1=580, c=4]
= 13.97
×
×
2nd quartiles, 𝑄 = 𝐿 + × 𝑐 [here = = 500occurred in 13-17]
= 13 + ×4 [ L1= 13, F1=110, f1=580, c=4]
= 15.69
×
×
3rd quartiles, 𝑄 = 𝐿 + × 𝑐 [here = = 750occurred in 18-22]
= 18 + ×4 [ L1= 18, F1=690, f1=175, c=4]
= 19.37
. × . .
We know, quartile coefficient of skewness = = = 0.36 Ans.
. .

H.W.Problem-2: Find the quartile coefficient of skewness for the wage distribution of the 65 employees:
Wages $250.00-$259.99 260.00-269.99 270.00-279.99 280.00-289.99 290.00-299.99 300.00-309.99310.00-319.99
No. of Employees 8 10 16 14 10 5 2
Ans: 0.0391