Statistics II

Class 7th April 2020

Measures of dispersion

Meaning of dispersion: Though average gives us one single value representing the entire
data, but it can’t be adequate in case all the observations are not same. It is thus necessary to
present the variability of the series since even though the average of the two given series may
be same but they may have wide variation. e.g.

Series A Series B Series c

100 100 1
100 105 2
100 102 5
100 101 9
100 92 483
Total :500 500 500
Average : 100 100 100

Since the average of three given series is 100 but the observations in three series is widely
different. Thus it is essential to understand the variations of items which is called dispersion.

Definition: “dispersion is the measure of variation of the items” (Bowley)

“Dispersion or spread is the degree of the scatter or variation of the variable about a central
value” (Brooks & Dick)

What are the significance of measuring dispersion?

1. To determine the reliability of an average: if dispersion is small, average is highly

representative of the series.
2. It serves as the base to control the variability: in case of quality control in production
units

1
 3. To compare two or more series with regard to their variability: higher variation little
uniformity.

Methods of measuring dispersion

1. Range
2. Interquartile Range and Quartile Deviation
3. Mean deviation/average deviation
4. Standard deviation

Absolute and relative measures of dispersion: absolute measures of dispersions are

expressed in the same unit in which original data is given. But if the two sets of data are
represented in two different units or their average size is very different or unmatchable e.g.
salary of professionals and labourers, relative measure of dispersion is used in such cases.

Measure of relative dispersion is the ratio of a measure of absolute dispersion to an

appropriate average. It is also called coefficient of dispersion.

I. Range

It is simplest method of studying dispersion. It is defined as difference between value of

largest and smallest item.

Range = L-S

L= Largest item, S= smallest item

Coefficient of range: L-S ∕ L+S

Exercise I. Calculate range of the price fluctuation in a given week. Also calculate
coefficient of range.

200, 150, 210, 190, 169, 250, 170

Solution :

2
 Range = L- S, where L= 250, S= 150

Thus range= 250-150 = 100

Coefficient of range: L – S / L+S

= 250 – 150 / 250+ 150

= 100/400 = 0.25

Exercise II. Calculate coefficient of range from the following data

Marks No. of students

10-20 8
20-30 10
30-40 12
40-50 6
50-60 3
60-70 2
70-80 9

(Hint: for continuous series, range is calculated as difference between the upper limit of
the upper class and lower limit of the lower class.

Coefficient of range: L –S / L+ S

Complete the exercise..............

Merits of Range:

1. Simplest of all the methods of dispersion and easy to compute

2. Take less time in computation so quick picture about dispersion can be obtained.

Uses of range:

3
 1. Weather forecast: minimum and maximum temperature
2. Share market fluctuations
3. Quality control
4. Everyday life, prices of items

Limitations of range:

1. It does not consider all the observation in the given series, only the two extreme items
2. It cannot tell us anything about the character of distribution
3. Cannot be calculated in case of open ended series.

II. Interquartile range and quartile deviation

Range which includes middle 50% of items in a distribution, one quarter of observations
at both the end are excluded in calculating inter-quartile range. Thus interquartile range
represents the difference between first quartile and third quartile.

Interquartile range : Q3 - Q 1

Quartile deviation: If interquartile range is reduced to half i.e. semi-interquartile range is

called as quartile deviation.

Quartile deviation (Q.D.) = Q3 – Q1 / 2

Q.D. = Q3-Q1
4
 2

If quartile deviation is very small it means high uniformity of central 50% of items.

Coefficient of quartile deviation: Q3 – Q1/ Q3 + Q1

Calculation of quartile deviation:

Individual series

Exercise III: 20, 28, 40,12,30,15,50

Solution :

Arrange the series in ascending or descending order

12, 15, 20, 28, 30, 40, 50

Q1 = size of N+1/4th item

= 7+1/4th item i.e. 2nd item = 15

Q3 = size of 3( N+1)/4th item

= size of 3 ( 7+1)/4th item

= size of 6th item i.e. 40

Q.D. = Q3-Q1
2
= 40-15/2

= 12.5

Coefficient of quartile deviation = Q3 - Q1/Q3 + Q1

= 40- 15/ 40+15

5
 = 25 / 55 = 0.455

Discrete series

Exercise IV : Calculate coefficient of quartile deviation from the following data:

Marks Frequency
10 5
20 7
30 16
40 6
50 10
60 1

Hint :

First Calculate C.f .

Calculate Q3, calculate Q1

Coefficient of quartile deviation : Q3-Q1/ Q3+ Q1

Solution

Marks Frequency C.f.

10 5 5
20 7 12
30 16 28
40 6 34
50 10 44
60 1 45
∑f=45

6
Q1= Size of N+1th item
4
= 45+1th item
4
11.5th item

Since 11.5th item lies in cf12, so Q1= 20

Q3 = Size of 3(N+1)th item

4
= size of 3(45+1)th item
4

= 34.5th item

Since 34.5th item lies in cf 44, so Q3= 50

Quartile deviation = Q3-Q1

2
Q.D.= 50-20
2
=15
Coefficient of Q.D.= Q3-Q1
Q3+Q1
= 50-20
50+20
=30/70
= 0.428

Continuous series

Quartile class:

For first quartile: size of N/4th item

For third quartile: size of 3N/4th item

𝑵
−𝒄.𝒇
Q1 = 𝑳+ 𝟒
×𝒊
𝒇

7
 𝟑𝑵
−𝒄.𝒇
Q3 = 𝑳+ 𝟒
×𝒊
𝒇

Quartile deviation = Q3 - Q1

Exercise V : Calculate quartile deviation and its coefficient from the given data

Marks Frequency
Less than 35 12
35 - 40 60
40 – 45 90
45 – 50 25
50 - 55 7
55 and above 6
Solution:

Marks Frequency c.f.

30- 35 12 12
35 - 40 60 72
40 – 45 90 162
45 – 50 25 187
50 - 55 7 194
55 -60 6 200
∑f=200

Hint :

First calculate cumulative frequency

Determine Q1 class by = size of N/4th item and determine quartile class through C.f.

𝑁
−𝑐.𝑓0
Calculate Q1 using the formula as given above : 𝐿 + 4
×𝑖
𝑓

Determine Q 3 class by = size of 3N/4th item and determine quartile class through c.f.

8
 3𝑁
−𝑐.𝑓0
Calculate Q3 using the formula as given above : 𝐿 + ×𝑖
4
𝑓

Calculate quartile deviation = Q3 – Q1

𝑄3−𝑄1
Calculate coefficient of quartile deviation = 𝑄3+𝑄1

Merit of quartile deviation

1. It is useful in case of open ended distribution

2. It is also useful in skewed distribution as 1/4th of items at both the ends are excluded.

Limitation

1. Since it ignores 50% of the items so can’t be regarded as adequate method of studying
dispersion.
2. Not capable of mathematical operations

References for practice questions:

https://www.mathsisfun.com/data/quartiles.html
https://www.wallstreetmojo.com/quartile-deviation/

9
 Class on 9th April
III. Mean deviation

Mean deviation is also known as average deviation. Mean deviation can be calculated both
from mean and median but theoretically there is an advantage in taking deviation from
median as sum of deviation of items from median ignoring sings is minimum.

|𝑫|
M.D = for individual series
𝑵

𝒇|𝑫|
Continuous series and discrete series = = 𝑵

𝑴.𝑫
Coefficient of M.D = 𝑴𝒆𝒅𝒊𝒂𝒏

(Note : if mean is used for calculation of mean deviation then coefficient of mean
deviation has to calculated from mean instead of median)

Where D = X-A ( A is average, Mean or median)

Steps :

 Calculate median of the series,

 Calculate deviation of items from median ignoring sings i.e |D|

10
  Obtain sum of |D| or go further to obtain f|D| in case of discrete and continuous series
 Divide the above sum from N

Exercise VII: Calculate mean deviation of the two given series and compare which
series has large mean deviation. Compare the coefficient of mean deviation of the two
series.

Series I : 400, 420, 440, 460, 480

Series II : 300, 400, 420, 440, 460, 480, 580

Hint:

 Calculate median of each series

 Calculate |D|= |X-median| for each series
|𝑫|
 Put the value in the formula M.D = for individual series
𝑵

Solution:

Series I Series II
X |D| = |X-A| X |D| = |X-A|
400 40 300 140
420 20 400 40
440 0 420 20
460 20 440 0
480 40 460 20
∑|D| = 120 480 40
580 140

∑|D| = 400
Series I

Median (A) = N+1 th item

2
= 5+1th item
2

= 3rd item, which is 440

11
 |𝐷|
Mean deviation:= 𝑁

M.D. = 120
5
M.D. = 24

Series II

Median (A) = N+1 th item

2
= 7+1th item
2

= 4th item, which is 440

|𝐷|
Mean deviation:= 𝑁

M.D. = 400
7
M.D. = 57.14

Coefficient of M.D.= M.D

Median
Coefficient of M.D for series I = 24
440
= 0.054

Coefficient of M.D. for series II = 57.14

440

= 0.129

Exercise VIII : calculate mean deviation from the given data

X F
10 3
11 12
12 18
13 12
14 3

Hint : calculate median using N+1/2th item

Calculate |D|, Calculate f|D|, compute ∑f|D|/median

12
Solution:

X f cf |D| = |X-A| f|D|

10 3 3 2 6
11 12 15 1 12
12 18 33 0 0
13 12 45 1 12
14 3 48 2 6
∑f= 48 ∑f|D| = 36

Median = N+1/2th item

Median = 48 +1
2

= 24.5th item

Since cf 33 has 24.5th item therefore the corresponding X value is the median = 12

Thus A= 12

𝑓|𝐷|
Mean deviation:= 𝑓

= 36
48
=0.75
Coffiecient of mean deviation = M.D.
Medain
= 0.75
12
= 0.0625 Answer

Exercise IX: Calculate mean deviation from the given data

X F
0-10 7
10-20 12
20-30 18
30-40 25
40-50 16

13
Hint :

𝒇|𝑫|
Mean Deviation = = 𝒇

|D| = |m-A|, m= mid point, A= median /mean

 Calculate median
 Calculate mid point and then calculate |D| = |m-A|
 Calculate f|D|
 Put the values in the above formula of mean deviation

X f cf m |D|=|m-A| f|D|
0-10 7 7 5 25.8 180.6
10-20 12 19 15 15.8 189.6
20-30 18 37 25 5.8 104.4
30-40 25 62 35 4.2 105
40-50 16 78 45 14.2 227.2
∑f=78 ∑f|D| =
806.8

Median class: = size of N/2th item

= 78/2 th item

= 39th item

Since 39.5th item lies in the cf 62, therefore the corresponding class is the median class

Median class = 30-40

𝑁
− 𝑐𝑓0
Median = 𝐿 + 2 𝑋𝑖
𝑓

78/2− 37
Median -= 30 + 𝑋 10
25

39− 37
-= 30 + 𝑋 10
25

30+20

14
 25
30+ 0.8

30.8 = A
𝒇|𝑫|
Mean Deviation (M.D.) = = 𝒇

806.8
= 𝟕𝟖

10.34 answer

Coefficient of mean deviation = M.D.

Median
= 10.34
30.8

= 0.335
Practice question

1. Calculate mean deviation and its coefficient from mean as well as from median.

X F
0-10 5
10-20 8
20-30 12
30-40 15
40-50 20
50-60 14
60-70 12
70-80 6

Practice question 2: calculate mean deviation from the following data.

Age Group Number of people

15-25 25
25-35 54
35-45 34
45-55 20

15
Sources for more practice questions:

https://www.topperlearning.com/cbse-class-11-science-mathematics/statistics/mean-deviation

https://www.mathsisfun.com/data/mean-deviation.html

https://www.analyzemath.com/statistics/mean-and-standard-deviation-problems.html

Merits

1. It is based on each and every item in the series

2. It is less affected by values of extreme items

Limitation

1. Since algebraic signs are ignored so not capable of further algebraic operations
2. Since median itself is positional average so mean deviation can’t give us accurate
results.

16
 Class on 13th April 2020
Standard deviation

Standard deviation is most important and widely used measure of studying dispersion. It
was introduced by Karl Pearson in 1823. It is also known as root mean square deviation.
It is denoted by the Greek letter σ (sigma). Greater the standard deviation, greater would
be the variability from the mean in the given series.

Advantages of standard deviation over mean deviation

1. In case of mean deviation algebraic signs are ignored while standard deviation
considers algebraic signs.
2. Mean deviation can be computed either from mean or median, but standard deviation
is strictly computed from mean.
3. It is best method of dispersion as it is governed by strict mathematical formula.
Individual series
Rule for the formula : root mean square deviation
𝑥2
1. Actual mean method σ = where x = X -X̅
𝑁

2. Assumed mean method

𝑑2 Ʃ𝑑
σ= − ( 𝑁 )2
𝑁

where d= X- A , A= assumed mean

Discrete series
𝑓𝑥 2
3. Actual mean method σ = where x = X -X̅
𝑓

4. Assumed mean method

17
 𝑓𝑑 2 Ʃ𝑓𝑑 2
σ= −( )
𝑓 𝑓

Continuous series
Step deviation method
𝑓𝑑 ´2 Ʃ𝑓𝑑 ´ 2
σ= −( ) Xi
𝑓 𝑓

Where d´= d/i, d= (m-A) , i = class interval

Exercise X: Calculate standard deviation from the following data

240,260,290,245,255,288,272,263,277,251

X d(X-A) A= 264 d2
240 -24 576
260 -4 16
290 26 676
245 -19 361
255 -9 81
288 24 576
272 8 64
263 -1 1
277 13 169
251 -13 169
∑X= 2641 ∑d= 1 ∑ d2 =2689
Solution:

Step 1: calculation of mean

𝑋
X̅ = 𝑁

= 2641
10
= 264.1
Since actual mean is in fractions, so we can take assumed mean for calculation of standard deviation

18
 𝑑2 Ʃ𝑑
σ= − ( 𝑁 )2
𝑁

d= (X-A), A is assumed mean

let A= 264

2689 1
σ= − (10 )2
10

σ = 268.9 − 0.01

σ = 268.89

σ = 16.398 Ans.

Calculation of standard deviation in discrete series

Exercise XI: Annual salary of a factory workers is given below. Calculate the standard
deviation of the salaries.

Salary (X) No. of persons(f)

45 3
50 5
55 8
60 7
65 9
70 7
75 4
80 7

Solution:

(X) (f) d=(X-A) d2 fd f d2

45 3 -15 225 -45 675
50 5 -10 100 -50 500
55 8 -5 25 -40 200
60 7 0 0 0 0

19
65 9 5 25 45 225
70 7 10 100 70 700
75 4 15 225 60 900
80 7 20 400 140 2800
∑f= 50 ∑fd=80 ∑f d2= 6000

Let us use assumed mean method

Let A = 60

𝑓𝑑 2 Ʃ𝑓𝑑 2
σ= 𝑓
−( 𝑓
)

6000 80 2
σ= 50
− 50

σ = 120 − 1.6 2

σ =√120-2.56
σ =√120-2.56
σ =√117.44
10.83 Answer

Exercise XII: Calculate standard deviation from the following:

Age f
0-10 15
10-20 15
20-30 23
30-40 22
40-50 25
50-60 10
60-70 5
70-80 10

20
Solution:

X f m d(m-A) d’=d/i d’2 fd’ f d’2

0-10 15 5 -40 -4 16 -60 240
10-20 15 15 -30 -3 9 -45 135
20-30 23 25 -20 -2 4 -46 92
30-40 22 35 -10 -1 1 -22 22
40-50 25 45 0 0 0 0 0
50-60 10 55 10 1 1 10 10
60-70 5 65 20 2 4 10 20
70-80 10 75 30 3 9 30 90
∑ f = 125 ∑ fd’= ∑ fd’2 =609
-123

𝑓𝑑 ´2 Ʃ𝑓𝑑 ´ 2
σ= −( ) Xi
𝑓 𝑓

Let A = 45

609 −123
σ= − ( 125 )2 X 10
125

σ = 4.87 − (−0.98)2 X 10

σ = 4.87 − (0.96) X 10
σ = 3.91 X 10

21
 = 1.97 X10
= 19.76 Answer

Merits of standard deviation

1. It included each and every item and governed by strict formula.

2. It is capable of further algebraic operation and further statistical measures like
correlation, less affected by fluctuations of sampling

Limitations

1. As compared to other measures, it involves difficult calculations

Practice question 1:

Calculate standard deviation of age from the given data

Age No. of persons

(X) (f)
20-25 170
25-30 110
30-35 80
35-40 45
40-45 40
45-50 35

Practice question 2

X f
3.5 3
4.5 7
5.5 22

22
 6.5 60
7.5 85
8.5 32
9.5 8

Relationship among various measures of dispersion

2 4
Q.D = 3 σ M.D = 5 σ

Class on 16th April 2020

Coefficient of variation

Standard deviation is absolute measure of dispersion, its relative measure is known as

coefficient of variation. It is calculated as :

σ
C.V. = X̅ x 100

Note: always use actual mean as average even if you are using assumed mean to calculate
standard deviation.

Coefficient of variation is used to compare the variation or consistency of two series.

Lower the value of coefficient of variation, higher the consistency of the series.

Step 1 : Find out the mean

Step 2: Find out the standard deviation

σ
Step 3: Use formula X̅ x 100

Step 4: Compare C.V. of the given series, the series with less CV will be more consistence.

23
Exercise XIII: Given below are the prices of two commodities over a period of ten days.
Find out which commodity is more stable.

Commodity X Commodity Y
30 110
59 105
50 107
54 103
57 108
59 105
51 107
52 101
48 103
50 101

Solution

Commodity X Commodity Y
2 2
Series 1 X
=
(X- x̄) X Series 2 X
=
(X- x̄) X

30 -21 441 110 5 25

59 9 81 105 0 0
50 -1 1 107 2 4
54 3 9 103 -2 4
57 6 36 108 3 9
59 8 64 105 0 0
51 0 0 107 2 4
52 1 1 101 -4 16
48 -3 9 103 -2 4
50 -1 1 101 -4 16
∑X=510 ∑ X 2= 643 ∑X=1050 ∑ X 2 = 82

24
Calculation for Series 1
𝑋
Mean = 𝑁
510
= 10

= 51
𝑥2
Standard deviation : σ = where x = X -X̅
𝑁

643
Standard deviation : σ = 10

σ = √64.3
8.02

σ
C.V. for series 1: = X̅ x 100
8.02
= x 100
51

15.72 for series 1

Calculation for Series 2

𝑋
Mean = 𝑁
1050
Mean = 10

= 105
𝑥2
Standard deviation : σ = where x = X -X̅
𝑁

82
Standard deviation : σ = 10

=2.86
σ
C.V. for series 2: = X̅ x 100
2.86
== x 100
105

=2.72 for series 2

Since coefficient of variation for series 2 is less than series 1. Therefore series 2 is
more consistent. Therefore price of commodity Y is more stable.

25
Exercise XIV: Following data shows marks of a class. Calculate average marks,
standard deviation and coefficient of variation of the marks.
Marks No. of students
20-30 8
30-40 14
40-50 12
50-60 18
60-70 13
70-80 9
80-90 6

Solution:
X f m d=(m-A) d2 fd fd2
20-30 8
30-40 14
40-50 12
50-60 18
60-70 13
70-80 9
80-90 6
∑f ∑fd ∑f d2

Practice question 1:

26
 Below are the marks scored by two students in a class Calculate coefficient of
variation and state which student is more consistent.
X Y
32 20
28 31
47 48
63 52
71 62
39 95
10 10
60 65
96 37
14 80

Source for more practice question

https://www.mathsisfun.com/data/standard-deviation.html

https://www.superprof.co.uk/resources/academic/maths/probability/normal-
distribution/standard-deviation-problems.html

https://www.examsolutions.net/tutorials/exam-questions-continuous-data-standard-deviation/

27