MAT2001-Statistics for Engineers

MEASURES OF CENTRAL TENDENCY

DESCRIPTIVE STATISTICAL MEASURES

Graphical representation summarizes information in the data. In addition to the diagrammatic

and graphic representations there are numerical methods which summarize the information in data sets,
called the measures of central tendency and dispersion. A measure of central tendency is a central value
around which the measurements have a tendency to cluster. Measures of dispersion estimate the extent
of the spread of the measurements of data sets. These measures put together are called descriptive
measures.

Whenever data are summarized, information on individual observations is lost. A measure of central
tendency can be computed for a sample or a finite population. The former is called statistic and the later
parameter. Sample mean is statistic while population mean is parameter.

MEASURES OF CENTRAL TENDENCY

A measure of central tendency is a typical value that serves as a representative of all the measurements.
The measurements obtained from a common source are not likely repetitions. It is undesirable to keep
all the measurements in focus. Consequently, a representative of all the measurements is required. Such
a representative is one of the three popular measures of central tendency. Viz., arithmetic mean, median
and mode. A measure of central tendency is a numerical value around which the measurements have a
tendency to cluster.
We come across five measures of central tendency (i) arithmetic mean, (ii) median, (iii) mode, (iv)
geometric mean and (v) harmonic mean.

Arithmetic Mean (or Simply Mean):

It is defined as the sum of the given observations divided by the total number of observations.
Sum of all observations
Arithmetic Mean (A.M.) = X =
Total number of observations
X =
X
n
where  X = Sum of all observations. (Read  as capital Sigma)
n = Total number of observations.
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
MAT2001-Statistics for Engineers

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Case A : Raw Data

Let X1, X2, …, Xn be ‘n’ measurements. The arithmetic mean of this data set can be computed by using
formula:

X=
 X , where  X =X1 + X2 + …,+Xn.
n
n = No. of observations in the given data.

Case B: Discrete frequency distribution

Consider the following discrete frequency distribution of variable values and their corresponding
frequencies

Variable Value (X) X1 X2 … Xk Total

Frequency (f) f1 f2 … fk N

The Arithmetic mean is then defined as,

X=
 fX , where  fX = f1X1+f2X2+ .. +fkXk
N
N = Total Frequency (  f )

Case C: Continuous frequency distribution

In this case, A.M. is given by X =  fX ,
N
where  fX = Sum of products of midvalues of class intervals and the
corresponding frequencies.
N = Total frequency. When mid values of class intervals are large in
magnitude, the step deviation method (or short cut method) can be employed to
find A.M.

Step deviation method : Under this method, A.M. can be calculated, using the following
formula :

X = A 
 fX 
 N  C
 

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

X A
where, d= = Scaled deviation of X.
C

X’s are mid values of class intervals.

A is a suitable origin parameter, usually chosen as the midpoint of the middle

most class interval.

C is the scale parameter whose value is the common width of the class
intervals.

fd is the sum of products of deviation values and their corresponding

frequencies.

Properties of Arithmetic mean:

(a) Sum of the deviations of observations taken from their A.M. is always zero. Symbolically, we
can write this property as
 (X - X ) = 0.
(b) The sum of the squares of the deviations of observations is minimum, when taken from A.M.
Symbolically, we write this as  (X - X )2 which is always minimum.

COMBINED ARITHMETIC MEAN

If X 1 and X 2 be arithmetic means of two series of n1 and n2 observations respectively, then the
arithmetic mean of combined data is as follows.
n1 X 1  n2 X 2
X= n1  n2

The combined arithmetic mean is based on n1  n2 observations.This formula can be generalized for
many number of groups say k
n X  n X ...  nk X k
X 1 1 2 2
n1  n2  ...  nk

Uses of Arithmetic Mean

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
It is one of the most commonly and widely used average. In the general usage, we talk about average
profits of a business concern, average production of an industry, average benefit of a group of persons
etc., we imply the arithmetic mean. Comparison of several means is an important problem in statistical
analysis.

Weighted Arithmetic Mean

In calculating arithmetic mean, we suppose that all the items in the distribution have equal importance.
But in practice, this may not be so. Some items may be more important than others. For example, rice,
wheat, oil, electricity, fuel, pulses etc are more important than coffee, tea, cigarettes, etc. Therefore,
proper weight age has to be given to various items in proportion to their relative importance for
calculating arithmetic mean. In such cases, weighted arithmetic mean is a suitable measure, for which
we assign different weights to different items according to their relative importance.
Suppose the weights assigned to the variable values (x1, x2, …, xn) be (w1, w2, …, wn) respectively.
Then the weighted arithmetic mean ( X w )is given by

Xw 
 WX
W
where WX = Sum of the products of variable values and their corresponding weights
W = Sum of weights.

SOLVED PROBLEMS ON ARITHAMETIC MEAN:

Problem: Calculate the mean height of the following 10 measurements

Height (in cms): 120, 115, 140, 141, 125, 124, 127, 130, 130, 133

Solution:
 X  1285
Number of measurements: n = 10

X
 X  1285  128.5
n 10
The mean height is 128.5 cms

Problem: Compute the arithmetic mean of daily wages of workers in a factory.

Worker : A B C D E F G H I J K L
Daily Wages:(in Rs.) 75 60 90 95 80 75 70 65 65 60 75 70

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Solution:
We have  X = 880
n = 12

X
 X  880  73.33
n 12
The arithmetic mean of daily wages of workers is Rs. 73.33.

Problem: The following data gives the number of children born to 350 women.

No. of children : 0 1 2 3 4 5 6
No. of women : 171 82 50 25 13 7 2
Calculate the mean number of children born per woman.

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Solution:
No. of children (x) No. of women (f) fx
0 171 0
1 82 82
2 50 100
3 25 75
4 13 52
5 7 35
6 2 12
Total 350 356
From the table, we have,
fx = 356 N = 350
X
 fx  356
N 350
X = 1.017
The mean no. of children born to a woman =1.017

Problem: Four teachers of Engineering reported grades of 80, 90, 50 and 60 respectively for 30,
40, 50 and 60 students, what is the average grade ?

Grade (x) No. of Students (f) fx

80 30 2400
90 40 3600
50 50 2500
60 60 3600
Total 180 12100
From table we have fx = 12100, N = 180

x
 fx  12100  67.22
N 180
Average Grade = 67.22

Problem: The following data relates to the marks of 100 students in statistics. Calculate the
A.M. marks of students.

Marks 20-30 30-40 40-50 50-60 60-70 70-80

No. of students 7 13 20 30 18 12

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Solution: (a) Direct Method :

Marks No. of students (f) Mid value (x) fx

20-30 7 25 175
30-40 13 35 455
40-50 20 45 900
50-60 30 55 1650
60-70 18 65 1170
70-80 12 75 900
Total 10 - 5250

From table we have, fx = 5250, N = 100

x
 fx  1520  52.5
N 100
Alternative Method:

(b) Step Deviation Method :

Let A = 55 and C = Length of class = 10

Marks No: of students Mid value x X A fd
d=
C
20-30 7 25 -3 -21
30-40 13 35 -2 -26
40-50 20 45 -1 -20
50-60 30 55 = A 0 0
60-70 18 65 1 18
70-80 12 75 2 24
Total 100 - - -25
From table we have fd = 25 , C = 10, A = 55, N = 100
  fd 
X  A    C
 N 
 25 
 55   10  55  2.5  52.5
 100 
Average Marks of Students = 52.5

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

Problem: compute arithmetic mean for the following frequency distribution:

Class : 50-59 60-69 70-79 80-89 90-99 100-109 110-119

Frequency 1 3 8 17 35 4 2
:

Solution: Given, C =Common length of class internals = 10

Class Frequency Mid value X A fd

(f) X d=
C
50-59 1 54.5 -3 -3
60-69 3 64.5 -2 -6
70-79 8 74.5 -1 -8
80-89 17 84.5 = A 0 0
90-99 35 94.5 1 35
100-109 4 104.5 2 8
110-119 2 114.5 3 6
Total 70 - - 32

From the table, we have, fd = 32, C = 10, A = 84.5, N = 70

  fd   32 
X  A    C  84.5   10
 N   70 
= 84.5 + 4.5714
Arithmetic Mean = X = 89.0714

Problem : The mean wage of workers in a factory running two shifts of 60 and 40 workers are
Rs.40 and Rs.35 respectively. Find the mean wages of all the 100 workers put
together.

Solution: Given that n1 = 60 and n2 = 40

X 1 = 40 and X 2 = 35

n1 X 1  n2 X 2
Combined A. M. = X =
n1  n2

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

(60)(40)  (40)(35)
=
60  40
2400  1400 3800
   38
100 100
Mean wage of all the 100 workers = Rs. 38.

Problem:Calculate the weighted A.M. of the index numbers for the following data:

Group Index No. Weight

Food 126 9
Clothing 130 5
Fuel and light 140 6
House Rent 175 2
Miscellaneous 182 3

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Solution:

Group Index No. (x) Weight (w) Wx

Food 126 9 1134
Clothing 130 5 650
Fuel and light 140 6 840
House Rent 175 2 350
Miscellaneous 182 3 546
Total - 25 3520

Weighted A.M = X w 
 wx  3520  140.8
 w 25
Weighted A.M. of index numbers is 140.8

Problem: Calculate the average salary paid in the whole industry, using the data given below:

Income Group (Rs.) 60-80 80-100 100-200 200-300 300-600

No. of Firms 16 13 12 10 14
Average No. of 4 7 5.25 2.2 1.5
workers

Solution : We are given frequency distribution with weights. We calculate weighted

arithmetic mean.

Total No. of
No. of firms Average No. of Mid value of
Class workers (weight) wx
(y) workers (z) class (x)
w = yz
60-80 16 4 70 64 4480
80-100 13 7 90 91 8190
100-200 12 5.25 150 63 9450
200-300 10 2.2 250 22 5500
300-600 14 1.5 450 21 9450
Total - - - 261 37070

Weighted Mean 
 wx  37070  142.0307
 w 261
Average salary paid to the workers in the industry is Rs. 142.03

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Problem: Given the following cumulative frequency distribution of less than upper bound
marks (x), obtained by 140 students in an examination, find the mean marks
of
students.

X 10 20 30 40 50 60 70 80 90 100
Cumulative
Frequency 140 133 118 100 75 45 25 9 2 0

Solution : Recover the frequency distribution from the given cumulative frequency
distribution constructing the following table

Lower More than Marks Class No. of students Mid value x fx

boundary cum. f Frequency (f)
10 140 10-20 7 15 105
20 133 20-30 15 25 375
30 118 30-40 18 35 630
40 100 40-50 25 45 1125
50 75 50-60 30 55 1650
60 45 60-70 20 65 1300
70 25 70-80 16 75 1200
80 9 80-90 7 85 595
90 2 90-100 2 95 190
100 0
Total - - 140 - 7170

From the table, we have

fx = 7170 N = 140
7170
X=  51.2143
140
Arithmetic mean marks of students = 51.2143

Problem: Find the missing frequency from the following data, given the average mark is
16.82

Marks Frequency

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
0-5 10
5-10 12
10-15 16
15-20 f4
20-25 14
25-30 10
30-35 8

Solution: Computation of Arithmetic mean :

Marks Midvalue Frequency (f) x  17.5 fd

X d=
5
0-5 2.5 10 -3 -30
5-10 7.5 12 -2 -24
10-15 12.5 16 -1 -16
15-20 17.5 = A f4 0 0
20-25 22.5 14 1 14
25-30 27.5 10 2 20
30-35 32.5 8 3 4
Total - N=70 + f4 - fd=-12
From the table, we have, N = 70 + f4, fd = -12 , C=5
  fd 
We know that, X -A+  C
 N 
 (12) 
16.82 = 17.5 +  5
 70  f 4 
60 60
16.82 - 17.5 =  0.68 
70  f 4 70  f 4
 0.68 (70+f4) = 60
 47.6 + 0.68 f4 = 60
 0.68 f4 = 60 – 47.6 =12.4
12.4
f4 =  18.2353 18
0.68

Problem: The mean wage of 100 workers in a factory who work in two shifts of 60 and 40
workers respectively is Rs. 38. The mean wage of 60 laborers working in the
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
morning shift is Rs, 40. Find the mean wage of laborers working in the
evening shift.

Solution : Given n1 = 60, n2 = 40

X1  40 and X  38
n X n X
we have, X = 1 1 2 2
n1  n2
(60)(40)  (40)( X 2 )
38 =
60  40
3800=2400+ X 2
40 X 2 =3800-2400=1400
1400
X2 = = 35
40
Mean wage of 40 workers working in the evening shift is Rs. 35.
Problem: The mean monthly salary paid to all employees in a certain company was Rs. 600/-,
the mean monthly salaries paid to male and female employees were Rs. 620
and 520 respectively. Obtain the percentage of male to female employees in
the company.

Solution : Mean salary paid to male and female employees X = 600

Mean salary paid to male employees X 1 = 620
Mean salary paid to female employees = X 2 = 520
n X n X 620n1  520n2
We have, X = 1 1 2 2 = 600=
n1  n2 n1  n2
600 (n1 + n2) = 620 n1 + 520 n2
600 n1 + 600 n2 = 620 n1 + 520 n2
600 n2 + 520 n2 = 620 n1 – 600 n1  80n2 = 20n1
n2 20 1
 
n1 80 4
 n1 : n2=4:1
4
Percentage of male employees =  100  80%
5
1
Percentage of Female Employees =  100  20%
5

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

Problem: For a certain frequency table, which has been partly reproduced here, the mean
was found to be 1.46.

No. of Accidents 0 1 2 3 4 5 Total

Frequency 46 ? ? 25 10 5 200

Find the missing frequencies.

Solution : Let f1 and f2 be missing frequencies, then

No. of accidents Frequency fx
0 46 0
1 f1 f1
2 f2 2f2
3 25 75
4 10 40
5 5 25
Total 86+f1+f2 140+f1+2f2
From table we have, f =N = (86 + f1 + f2)
fx = (140 + f1 + 2f2)

But given that

f = N = 200

Also X =
 fx  1.46
N
fx = N X = (200) (1.46) = 292
 We have 86 + f1 + f2 = 200 (1)
and 140+ f1 + 2f2 = 292 (2)

From (1) : f1 + f2 = 200 – 86 = 114 (3)

From (2) : f1 + 2f2 = 292 – 140 = 152 (4)

Solving (3) and (4), we get f1 and f2.

Consider (3) : f1 + f2 = 114
(4) : f1 + 2f2 = 152
------------------
By subtraction - f2 = - 38 or f2 = 38
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
------------------

By substituting f2 = 38 in (3)
we get, f1 + 38 = 114
 f1 = 114 – 38 = 76
Hence, the missing frequencies are f1 = 76 and f2 = 38

Problem: The mean salary paid to 1,000 employees of an establishment was found to be Rs.
180.40. Later on, after disbursement of salary, it was discovered that the salary
of two employees were wrongly entered as Rs. 297 and Rs. 165. Their correct
salaries were Rs. 197 and 185. Find the true arithmetic mean.

Solution: Given that X = 180.40 and n = 1000

Since X =
 X or  X  nX  (1000)(180.40) = 180400
n
Since, two values were wrongly entered, consider the correction as :

Corrected X = Wrong X – (Sum of wrong values) + (Sum of correct values)

= 180400 – (297 + 165) + (167 + 185)

= 180400 – 462 + 382
= 180320

Corrected  X 180320
Corrected Mean =   180.32
n 1000
True mean is Rs.180.32.

Problem: The mean weight of a student in a group of six students is 119 pounds. The
individual weights of five of them are 115, 109, 129, 117 and 114 pounds. What
is
the weight of the sixth student.

Solution : Given that X = 119 and n = 6

X
Since X = or  X  nX , we get,  X = (6) (119) = 714
n
 X = 714 = Sum of six observations
The sum of given five observations = 115 + 109 + 129 + 117 + 114 = 584
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
The value of sixth observation = Sum of 6 observations – Sum of five
observations.
= 714 – 584 = 130
The weight of sixth student = 130 pounds

Median

Median of data is the value of the variable which divides arranged data into two equal
parts. When the observations in the data are arranged either in ascending (increasing)
order or in descending (decreasing) order, the Median is defined as the value of the
variable, which divides the arranged data into two equal parts. We denote it by X

Case A : Raw data : When the number of observations in the data is an ‘odd’ number, then the
Median is equal to the middle most term in the arranged data. If the
number of observation is even, the simple average of the two middle most
values of the arranged data is median.

Case B :
Discrete frequency distribution : In this case, median is defined as the value of the variable
that
corresponds the less than cumulative frequency just above
half the total frequency.

Case C :
Continuous frequency distribution : In the case of continuous frequency distribution, median
can be obtained by using the following formula :

N 
 2 m
Median = X = L   C
 f 
 
where, L = Lower boundary of the median class
f = Frequency of the median class
m = Less than cumulative frequency of the class that precedes the
median
class
C = Length of the median class
N = Total Frequency

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

Here, Median class is the class that corresponds the less than cumulative
frequency just greater than (N/2) value.

Remark : When, we are given the class intervals in descending order, the median can be
found
by using the following formula.

N 
 2 m
Median = X = U -  C
 f 
 
where, ‘U’ is the upper bound of the median class. Other symbols such as N, m,
f and C remain the same as in the previous formula.

USES OF MEDIAN

It is used frequently in practice. It is readily used in cases where quantitative

measurement of all items is difficult but ordering of items is relatively easy. When the
data contains extreme values, it is advantageous to use median. Its usefulness as a
positional average is recognised in statistical analysis.

Remarks :

1. For a continuous frequency distribution Median can be found by plotting

gives.

2. The sum of the absolute deviations of observations taken from the median is
always minimum, i.e., X- X    | X  A | , where A is any arbitrary value.

SOLVED PROBLEMS ON MEDIAN

Problem 2.10.1: Determine Median for the following data : 26, 20, 15, 45, 18, 8, 10, 38, 13.

Solution : Arrange the observations in ascending order as

8, 10, 13, 15, 18, 20, 26, 38, 45
Number of observations = 9 (odd number)
Value of the middle term = Median = 18.
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

Problem 2.10.2: Find the median for the following values :

16, 12, 5, 8, 9, 5, 10, 28.

Solution : Arrange the observations in ascending order as,

5, 5, 8, 9, 10, 12, 16, 28.
Number of observations = 8 (even number)
Median = Average of values of two middle terms

9  10
Median =  9.5
2

Problem 2.10.3: Calculate the median for the following frequency distribution.

Variable X 3 4 5 6 7 8
Frequency (f) 2 4 7 9 10 16

Solution :
X f Less than cum. f
3 2 2
4 4 6
5 7 13
6 9 22
7 Median 10 32 Median N/2 = 24
8 16 48

Total N = 48 -

N 48
From the above table,   24
2 2
N
Here, ‘32’ is just greater than   value.
2
Median = The value of the variable that corresponds less than cumulative
frequency just greater than N/2 value = 7.

Problem: The weight of 30 students are given in the following frequency distribution.
Determine the median weight of the students.
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

Weight in kgs 60-64 65-69 70-74 75-79 80-84

No. of students 2 8 12 5 3
Solution :

Class Boundaries F Less than cum. f

59.5-64.5 2 2
64.5-69.5 8 10 = m N/2 = 15
69.5-74.5 Median class 12 =f 22
74.5-79.5 5 27
79.5-84.5 3 30
Total N = 30 -

N 30
From table,   15 Here, 69.5 – 74.5 is the median class.
2 2
We have, = 69.5; m = 10; f = 12; and C = 5
 N m
 
Median  U   2 C
 f 
 
15  10  25
 69.5    5  69.5   69.5  2.08  71.58
 12  12
Median weight = 71.58 kg.

Problem: Find the median weight for the following table:

Weight in kgs 84-80 79-75 74-70 69-65 64-60

No. of students 3 5 12 8 2

Solution: The class intervals are given in the descending order:

Class Boundaries f Less than cum. f

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
84.5-79.5 3 3
79.5-74.5 5 8=m
74.5-69.5 Median class 12 = f 20
69.5-64.5 8 28
64.5-59.5 2 30
Total 30 -
Here 74.5 – 69.5 is the median class.
N 30
From table ,   15
2 2
U = 74.5; m = 8, f = 12, and C = 15
N 
 2 m
Median = X = U -  C
 f 
 
15  8  35
 74.5    5  74.5 
 12  12
 74.5  2.92  71.58
Median weight = 71.58 kg.

Problem: In the frequency distribution of 170 families given below, the number of families
corresponding to expenditure groups 20-30 and 40-50 are missing in the table.
However, the median is known to be 35. Find the missing frequencies.

Expenditure 0-10 10-20 20-30 30-40 40-50 50-60 60-70

No. of families 10 20 ? 40 ? 25 15

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Solution : Let the missing frequencies of the expenditure groups 20-30 and 40-50 be f1 and f2
respectively.
Given , Median = 35 and total frequency = 170

Class f Less than cum. f

0-10 10 10
10-20 20 30
20-30 f1 30 + f1 = m
30-40 Median class 40 = f 70 + f1
40-50 f2 70 + f1 + f2
50-60 25 95 + f1 + f2
60-70 15 110 + f1 + f2
Total 170 -
Since, median = 35, the median class is given by 30-40.
From the table, we have
L = 30, f = 40, m = 30 + f1
N 170
  85 and C  10
2 2
N 
 2 m
Median = L +  C
 f 
 
 85  (30  f1 ) 
 35 = 30 +   10
 40
55  f1
 35  30 
4
 55  f1  20 or f1  55  20  35
Since, the total frequency N = 170, or 110 + f1+f2 = 170, and f1 = 35, we get
110 + 35 + f2 = 170 or f2 = 170 – 145 = 25
The missing frequencies are respectively f1 = 35 and f2 = 25.

MODE

Mode is the value of the variable, which occurs most frequently in the measurements.
The word ‘Mode’ is derived from the French word “La Mode’ which signifies fashion. It
is often used as a positional average in practice. Mode need not be unique. A data set

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
may have two or more modes. If a data set contains two modes then the data is said to be
bi-model data. Mode is usually denoted by X.

Case A :
Raw data : In this case, mode is defined as the value of the variable which occurs most
frequently in the given data set.

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Case B :
Discrete frequency distribution : In the case of discrete frequency distribution, mode is
the value of the variable for which the frequency is maximum.

Case C :
Continuous frequency distribution : In the case of continuous frequency distribution,
mode can be obtained by using the following formula.

 f  f1 
Mode = X = L +  C
 2 f  f1  f 2 
where L : Lower bound of the model class
f : Frequency of the model class
f1 : Frequency of the class that precedes the model class
f2 : Frequency of the class that succeeds model class
C : Length of the model class

Here, model class is the class that corresponds largest frequency.

Remarks :
In the following three cases, mode can not be obtained by using the above
formula:
(a) When the highest frequency is observed at the beginning of the frequency
table.
(b) When the highest frequency is observed at the ending of the frequency table.
(c) When two or more class intervals contain the same maximum frequencies.

However, in the above three cases, mode can be obtained by using either a method called
‘Grouping method’ or ‘empirical relationship between arithmetic mean, median and
mode.

Empirical Relationship between A.M. Median and Mode

If mean, median and mode are equal for a frequency distribution, then the distribution is
called symmetrical distribution. In a moderately asymmetrical distribution, mean,
median and mode approximately satisfy the following empirical relationship:

(A.M. - Mode) = 3 (A.M. - Median)

Symbolically ( X  X ) = 3 ( X  X )
Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

This relation is observed from experience and it is not mathematically derived. When we
are given the values of any two of these three measures, then the third measure can be
found from this relation. This relation is occasionally used to find mode, when A.M. and
Median are known.

Mode = 3 Median – 2 A.M.

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Uses of Mode :

Mode finds an important place in marketing studies, where a manager of a business

concern is interested in knowing about the size which has the highest concentration of
items. For example, in placing an order for shoes or ready made garments, the model
size helps because this size and other sizes around it are in common demand. It is also
used in dealing with non-quantitative data.

SOLVED PROBLEMS ON MODE

Problem: Find the mode for the following set of values of a variable :
2, 7, 3, 2, 1, 3, 2, 2, 5

Solution : In the given data the value 2 occurs most frequently than the other values.
Hence,
mode of the given data is 2.

Problem: Determine the mode for the following data

1, 0, 2, 3, 6, 7, 5, 4, 8.

Solution : In the given data, no single value repeats more frequently, when compared to
other
values. Therefore, we conclude that there is no mode for the above data.

Problem: Find the modal age of married women at first child birth :

Age at the birth of first 13 14 15 16 17 18 19 20 21 22 23

child
(in years)
No. of married women 37 150 300 360 270 435 160 200 85 65 25

Solution : From the given discrete frequency distribution, it is observed that the highest
frequency
is 435. The age of married women corresponding to this highest frequency is 18
years.
Hence, 18 years is the model age of married women at first child birth.

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Problem Calculate the mode for the following data :

Class 130-134 135-139 140-144 145-149 150-154 155-159 160-164

Frequency 5 15 28 24 17 10 1

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
Solution :

Class Frequency
130-134 5
135-139 15 = f1
140-144 Model class 28 = f
145-149 24 = f2
150-154 17
155-159 10
160-164 1
Total 100

The highest frequency corresponds the class 140 –144. Therefore, it is the
model class.

From table, we have,

L = 139.5, f = 28, f1 = 15, f2 = 24 and C = 5

 f  f1 
Mode = X = L +  C
 2 f  f1  f 2 
 28  15 
 139.5   5
 2 x 28  15  24 
65
 139.5   139.5  3.8235
17
 143.3235

Problem: In a moderately skewed distribution (Asymmetrical distribution) A.M. = 15 and

Mode = 12. Find the value of the Median of the given distribution.

Solution : Consider the empirical relationship between mean, median and mode.

(Mean - Mode) = 3 (Mean - Median)

Hence, 15 - 12 = 3 (15 - Median)

Therefore, 3 Median = 45 – 3 = 42

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

42
Median =  14
3

Problem: Find the Mode for the following data :

2, 5, 3, 2, 1, 4, 6, 3, 7

Solution : Since 2 and 3 have maximum frequencies, (2 occurs 2 tm ice 3 occurs 2 tm ice)
the
given data is a bimodal data. Therefore, we use the empirical relationship
between
mean, median and mode.
 x 33
A.M  x    3.6667
n 9

To find median arrange the data in ascending order:

1, 2, 2, 3, 3, 4, 5, 6, 7
Median = 3
Consider the empirical relationship (AM. - Mode) = 3 (Mean - Median)
Substituting the values of A.M. and Median. we get.

3.6667 – Mode = 3 (3.667-3)

3.6667 – Mode = 2.0001
Mode = 3.667 – 2.0001 = 1.6666

Problem: Compute Mode for the following data :

Size 0-4 4-8 8-12 12-16 16-20

Frequency 10 20 30 35 35

Solution: The highest frequency is 35 and it corresponds the two bottom most class intervals
of
frequency table. Hence, the given distributions is a bimodal. In this case, we use
the
empirical relationship between A.M., Median and Mode.

Class Frequency f Mid values x fx Less than cumulative

frequency

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
0-4 10 2 20 10
4-8 20 6 120 30
8-12 30 10 300 60 =m
12-16 35 = f 14 490 95
16-20 35 18 630 130
Total 130 - 1560 -

Form the table,  fx = 1560; N = 130

 fx 1560
A.M. = x    12
N 130
Since, N/2 = 130/2 = 65
12-16 is the median class
L=12, m = 60, f = 35, N = 130, C = 4
N 
 2 m
Median = L +  C
 f 
 
 65  60  20
 12    4  12   12  0.5714  12.5714
 35  35
Median = 12.5714
Consider the empirical relationship
(A.M. - Mode) = 3 (Mean - Median)
12 – Mode = 3 (12 – 12.5714)
12 – Mode = 3 (- 0.5714)
12 – Mode = - 1.7142
Mode = 12 + 1.7142
Mode = 13.7142

Problem: The median and mode of the following wage distribution are known to be Rs.
335 and Rs. 340 respectively. Find the three missing frequencies in the frequency
distribution given below.

Wages (in Rs.) Frequency

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers
0-100 4
100-200 16
200-300 60
300-400 ?
400-500 ?
500-600 ?
600-700 4
Total 230

Solution : Let f1, f2 and f3 be the missing frequencies.

Wages ( Rs.) Frequency Less than cumul-

ative frequence
0-100 4 4
100-200 16 20
200-300 60 80
300-400 Median class f1 80 + f1
400-500 f2 80 + f1 + f2
500-600 f3 80 + f1 + f2 + f3
600-700 4 84 + f1 +f2 + f3
Total 230
We have, 230 = 84 + f1 + f2 + f3
f3 = 230 – 84 – f1 – f2
= 146 – (f1 + f2)………..(1)
Since the median and mode are 355 and 340, both lie in the same class, viz., 300 –
400.

 230 
 2  80 
Median = 335 = 300 +   100
 f1 
 
(115  80)100
335-300 =  35 f1  3500
f1
3500
 f1   100
35

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

 100  60 
Mode = L +   100
 2  100  60  f 2 
4, 000
40   40(140  f 2 )  4000
140  f 2
 5600  40 f2  4000
40 f2  5600  4000  1600
1600
 f2   40
40
From (1), we get f3 = 146 - (100 + 46) = 6
Hence, the missing frequencies are 100, 40 and 6 respectively

Requisites for an Ideal Measure of Central Tendency :

According to Prof. Yule, the following are the chief characteristics to be satisfied by an
ideal measure of central tendency.
i) It should be rigidly defined
ii) It should be easy to understand and easy to calculate
iii) It should be based on all the observations.
iv) It should be suitable for further mathematical manipulations
v) It should be affected as little as possible by sampling fluctuations.
vi) It should not be affected much by extreme values.

Now, we shall consider the merits and demerits (or advantages and disadvantages) of
Arithmetic mean, median and mode.

Dr.Mokesh Rayalu,M.Sc,Ph.D.,
 MAT2001-Statistics for Engineers

Measure of Central Merits Demerits :

Tendency
a) Arithmetic Mean 1) It is rigidly defined. 1) It is affected very
2) It is easy to understand and much by extreme
easy to calculate. values
3) It is based on all the 2) It cannot be found for
observations. distributions with
4) It is suitable for further open end classes.
mathematical 3) It cannot be used if
manipulations. measurements are
5) Of all the averages, it is qualitative.
affected least by sampling
fluctuations.

1) It is rigidly defined.
b) Median 2) It is easy understand and 1) It is not based on all
easy to calculate. the observations.
3) It is not affected by 2) It is not suitable for
extreme values. further mathematical
4) It can be found for manipulations.
distributions with open end 3) It is affected by
classes. sampling fluctuations.
5) Median is the only average 4) It is not so generally
to be used while dealing familiar as the
with the qualitative data. arithmetic mean.

1) It is easy to understand and

easy to calculate.
c) Mode 2) It is not affected by
extreme classes.
3) It can be found for 1) Mode is ill-defined
distribution with open end 2) It is not based on all
classes. the observations.
3) It is not suitable for
further mathematical
manipulations.
.
Dr.Mokesh Rayalu,M.Sc,Ph.D.,