Bcom IIIrd Semester

Basic
Numerical
Methods
University of Calicut
 CHAPTER-V
DESCRIPTIVE STATISTICS

Basic Numerical Methods

Measures of Dispersion
The degree to which numerical data tend to spread about an average value is
called dispersion. If the observation are very close to the center, we say that
dispersion is small. If the observations are spread away from the center, we say
dispersion is large.

Absolute Dispersion:
Dispersion can be expressed either in terms of original units or an
abstract figure like ratios or percentages . If the dispersion expressed in
terms of its original units are called absolute dispersion
Relative Dispersion
If the dispersion measured as ratio or percentages of average is called
Relative measure of dispersion.
The following Methods are used to measure the absolute dispersion and
relative measure.

Absolute dispersion Relative dispersion

Range Coefficient of Range

Quartile Deviation Coefficient of QD

Mean Deviation Coefficient of MD

Standard Deviation Coefficient of variation

 Range

The difference between the largest value(L) and smallest value(S) in

a series called the range of the distribution.

Absolute measure of Range = L-S

𝐿−𝑆 𝑊ℎ𝑒𝑟𝑒 𝐿 − 𝑙𝑎𝑟𝑔𝑒𝑠𝑡 𝑖𝑡𝑒𝑚 &
Relative measure of range -Coefficient of Range =
𝐿+𝑆 𝑆 − 𝑆𝑚𝑎𝑙𝑙𝑒𝑠𝑡 𝑖𝑡𝑒𝑚

If the range is higher implies higher dispersion in the series and vice versa
 Merits Demerits
1. It is the simplest measure of 1. It gives importance to the two
dispersion. extreme values only, and therefore
2. It is simple to understand and easy to it may be unduly influenced by
calculate. extreme values.
2. It is not a reliable measure of
dispersion on many occasions.

Example-1:
Series A:- 1400,1450,1520,1380,1485,1495,1575,1440 so the range of this
series is given by
R=1575-1380=195 and

1575−1380 195
coefficient of range = = = 0.0659
1575+1380 2955
 Quartile Deviation

The difference between 3rd and 1st quartile, called inter quartile range. and
half of the difference between 3rd and 1st quartile is called quartile
deviation(or semi inter quartile range)
𝑄3 −𝑄1
Absolute measure :Q.D=
2 𝑊ℎ𝑒𝑟𝑒 𝑄1 − 𝐹𝑖𝑟𝑠𝑡 𝑄𝑢𝑎𝑟𝑡𝑖𝑙𝑒 &
𝑄3 − 𝑡ℎ𝑖𝑟𝑑 𝑞𝑢𝑎𝑟𝑡𝑖𝑙𝑒
𝑄3 −𝑄1
Relative measure :Coefficient of Q.D =
𝑄3 +𝑄1
Example-2: Calculate the Quartile deviation and coefficient of Quartile deviation
maximum load no.of cables N 60
9.3-9.7 2 = = 15
9.8-10.2 5
4 4
10.3-10.7 12 𝑙 = 10.25
10.8-11.2 17 cf= 7
11.3-11.7 14
11.8-12.2 6
f= 12
12.3-12.7 3 C= 0.5
12.8-13.2 1
Solution:
maximum load no.of cables class boundary cf
9.3-9.7 2 9.25-9.75 2
9.8-10.2 5 9.75-10.25 7
10.3-10.7 12 10.25-10.75 19
10.8-11.2 17 10.75-11.25 36
11.3-11.7 14 11.25-11.75 50
11.8-12.2 6 11.75-12.25 56
12.3-12.7 3 12.25-12.75 59
12.8-13.2 1 12.75-13.25 60
 N
−cf
Q1 = l + 4
×𝐶 N 60
f = = 15
15 − 7 4 4
Q1 = 10.25 + × 0.5 𝑙 = 10.25
12
=10.58 cf= 7
f= 12
C= 0.5
3N
− cf
Q3 = l + 4 ×𝐶
f
N 3 × 60
3 = = 45
4 4
45 − 36 𝑙 = 11.25
Q3 = 11.25 + × 0.5
14 cf= 36
f= 14
Q3 = 11.25 + 0.32 = 11.57 C= 0.5
 𝑄3 −𝑄1 11.57−10.58
Q.D= = 0.495
2 2

𝑄3 −𝑄1 11.57−10.58
Coefficient of Q.D = = = 0.044
𝑄3 +𝑄1 11.57+10.58

Merits Demerits
1. It can be calculated for open end 1. It is not based on all
classes. observations.
2. It is not unduly affected by 2. It is not capable of further
extreme values. algebraic treatments.
 Mean Deviation about mean

Mean deviation is the arithmetic mean of absolute deviation of observation

from their average , the average may be mean , median or mode .

Mean Deviation about mean Mean Deviation about mean

(Individual series) (Discrete &Continuous series)
𝑓|𝑥−𝑥|
|𝑥−𝑥| M.D=
𝑁
M.D= 𝑀𝐷
𝑁 Coefficient of MD=
𝑀𝐷 𝑚𝑒𝑎𝑛
Coefficient of MD=
𝑚𝑒𝑎𝑛
 Mean Deviation about median

Mean Deviation about median Mean Deviation about median

(Individual series) (Discrete &Continuous series)
𝑓|𝑥−𝑀|
M.D=
|𝑥−𝑀| 𝑁
M.D=
𝑁
𝑀𝐷
Coefficient of MD=
𝑀𝐷 𝑀𝑒𝑑𝑖𝑎𝑛
Coefficient of MD=
𝑀𝑒𝑑𝑖𝑎𝑛
 Mean Deviation about mode

Mean Deviation about mode Mean Deviation about mode

(Individual series) (Discrete &Continuous series)
𝑓|𝑥−𝑀𝑜𝑑𝑒|
M.D=
|𝑥−𝑀𝑜𝑑𝑒| 𝑁
M.D=
𝑁
𝑀𝐷
Coefficient of MD=
𝑀𝐷 𝑀𝑜𝑑𝑒
Coefficient of MD=
𝑀𝑜𝑑𝑒
Example-3: Calculate the mean deviation from (1)mean (2)median
(3)mode in respect of the marks obtained by nine students
gives below and show that mean deviation from median is
minimum. marks: 7,4,10,9,15,12,7,9,7
Solution:
Arranging the observations in ascending order
Marks; 4,7,7,7,9,9,10,12,15

80
𝑚𝑒𝑎𝑛(𝑋) = = 8.89
9
𝑛+1 𝑡ℎ 9+1
Median= 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑖𝑡𝑒𝑚 = 5𝑡ℎ 𝑖𝑡𝑒𝑚 = 9
2 2

Mode=7(most occurring observation)

 X |X-𝑋 | |X-M| |X-Mode| 21
4 4.89 5 3 M.D (about median) = = 2.33
9
7 1.89 2 0
7 1.89 2 0 2.33
7 1.89 2 0 Coefficient of MD (about median)= = 0.258
9 0.11 0 2 9
9 0.11 0 2
10 1.11 1 3 23
12 3.11 3 5
M.D (about mode) = = 2.56
9
15 6.11 6 8
80 21.11 21 23 2.56
Coefficient of MD (about mode)= = 0.365
7
21.11
M.D (about mean) = = 2.35
9
2.35
Coefficient of MD (about mean)= = 0.264
8.89
 Merits Demerits
1. It is based on all observations. 1. Ignoring signs of deviations
2. It can be calculated from any may create artificiality.
value. 2. Computation is not much
3. It is not much affected by easier.
extreme values.
 Standard Deviation(𝜎)

Standard deviation was introduced by karl pearson ; standard deviation is

defined as the positive square root of the mean on of squares of deviation
from arithmetic mean . It is denoted as Greek letter "𝜎” (sigma).
If for a distribution all values are equal standard deviation will be zero ;so
that the minimum value of standard deviation is zero. So the best
measure of dispersion is standard deviation.
The sum of squared deviation is minimum when taken about mean.

Standard Deviation(Individual Series)

(𝑥−𝑥)2 𝑥2 𝑥 2
Actual mean method 𝜎 = = −( )
𝑁 𝑁 𝑁

𝑑2 𝑑 2
short cut Method 𝜎= −( ) where
𝑁 𝑁
d=X-A
 Standard Deviation(𝜎)

Standard Deviation(Discrete and Continuous Series)

𝑓(𝑥−𝑥)2 𝑓𝑥2 𝑓𝑥 2
Actual mean method 𝜎 = = −( )
𝑁 𝑁 𝑁

𝑓𝑑2 𝑓𝑑 2
short cut Method 𝜎= −( ) where
𝑁 𝑁
d=X-A
𝑓𝑑′2 𝑓𝑑′ 2
Step Deviation 𝜎= −( ) × 𝑐 where 𝑑 =
𝑁 𝑁
𝑋−𝐴
𝑐
 v𝐚𝐫𝐢𝐚𝐧𝐜𝐞(𝜎 2 )

variance = 𝜎 2 𝑜𝑟 𝑆𝐷 = variance

Properties of variance
Variance and consequently standard deviation is independent of change of origin.

For any discrete frequency distribution Standard Deviation is not less than mean
deviation ie; SD>MD from mean
 Coefficient of variation(CV)

𝜎
Coefficient of variation or CV= × 100
𝑋

For comparing two series we calculate CV , For each series . The series having
greater CV is said to be more variable than the other, and the series having
lesser
CV is said to be more consistent or stable.
 Standard Deviation(𝜎)

Merits Demerits
1. It is rigidly defined and based on 1. It is difficult to calculate.
all observations. 2. It is impossible to find it in
2. It is capable of further algebraic open end classes.
treatments.
3. It is less affected by the
fluctuations of sampling
than other measures of dispersion.
Example-4: Calculate standard deviation from the following distribution of
marks by using all methods
marks 1-3 3-5 5-7 7-9
No. of Students 40 30 20 10

Solution: Actual mean method

mark f X fX 𝑋−𝑋 2 𝑓 𝑋−𝑋 2

1-3 40 2 80 4 160
3-5 30 4 120 0 0
5-7 20 6 120 4 80
7-9 10 8 80 16 160
100 400 400
𝑓(𝑥−𝑥)2 400
Actual mean method 𝜎 = = 4=2
𝑁 100
Assumed mean method
mark f X d=(X-2) fd 𝑓𝑑 2
1-3 40 2 0 0 0
3-5 30 4 2 60 120
5-7 20 6 4 80 320
7-9 10 8 6 60 160
100 200 800

𝑓𝑑 2 𝑓𝑑 2 800 200 2
short cut Method 𝜎= − − = 8−2= 4=2
𝑁 𝑁 100 100
Step deviation method

′ 𝑋−2
mark f X 𝑑 = 𝑑′
2
𝑓𝑑′ 𝑓𝑑′
2
2
1-3 40 2 0 0 0 0
3-5 30 4 1 1 30 30
5-7 20 6 2 4 40 80
7-9 10 8 3 9 30 90
100 100 200

𝑓𝑑 ′2 𝑓𝑑′ 2 200 100 2

Step Deviation 𝜎= −( ) × 𝑐 − ×2
𝑁 𝑁 100 100

= 1 ×2

=2