[go: up one dir, main page]

0% found this document useful (0 votes)
370 views15 pages

Chapter 6

This document discusses measures of central tendency, including the arithmetic mean. It defines the arithmetic mean as the sum of all values divided by the number of values. The key properties discussed are: 1) The arithmetic mean is the central value about which positive and negative deviations from other values balance out, with their sum equaling zero. 2) For the arithmetic mean to be a satisfactory measure of central tendency, it should be based on all values, easy to calculate and interpret, and amenable to algebraic treatment. 3) A single very large or small value can unduly influence the arithmetic mean, since it is calculated using all values.

Uploaded by

Abhinav Mohanty
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
370 views15 pages

Chapter 6

This document discusses measures of central tendency, including the arithmetic mean. It defines the arithmetic mean as the sum of all values divided by the number of values. The key properties discussed are: 1) The arithmetic mean is the central value about which positive and negative deviations from other values balance out, with their sum equaling zero. 2) For the arithmetic mean to be a satisfactory measure of central tendency, it should be based on all values, easy to calculate and interpret, and amenable to algebraic treatment. 3) A single very large or small value can unduly influence the arithmetic mean, since it is calculated using all values.

Uploaded by

Abhinav Mohanty
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

48 INTRODUCTION TO STATISTICS

CHAPTER 6

Measures of Central Tendency

We have discussed classification and 1. Averages


tabulation of data and their
The term ‘average value’ is used by us in
diagrammatic presentation in the earlier
every day life. It is that ‘single value’
chapters. They help in visual comparison
which represents all values in a given set
and interpretation of data and facilitate
of values. For example, the average height
drawing conclusions about the salient
of students in your class represents
features of the population.
heights of all students in the class. The
However, we should like to know
average income of people in a certain
more about the distribution of values of
region represents income of all people in
the variable on which the data have been
that region.
collected. For example,
The following averages are suitable in
(a) what are the largest and the smallest
different situations :
values of the variable, and what is the
(i) arithmetic mean
range of the distribution of values,
(ii) median
(b) what is the central value of the
(iii) mode
distribution about which other values
(iv) geometric mean
are distributed,
(v) harmonic mean.
(c) what is the amount of variation in
Before we discuss their properties, let
values about the central value, and
us discuss the requirements of a
(d) are the values of the variable
satisfactory average. (We omit discussion
uniformly distributed over the range,
of geometric and harmonic means as they
or are most of them clustered about
are not in the syllabus for Class XI
some value in the upper or lower part
students.)
of the range, or in the middle of the
range?
2. Requirements of a Satisfactory
In this chapter, we propose to discuss
Average
the measures of central tendency and
postpone discussion of measures of A satisfactory average should be
dispersion to the next chapter. (i) rigidly defined
The measures of central tendency (ii) based on all values
are also called measures of location, or, (iii) simple and easy to interpret
briefly averages. (iv) easy to calculate
MEASURES OF CENTRAL TENDENCY 49

(v) amenable to algebraic treatment TABLE 6.1


If the average is not rigidly defined, it Deviations of Marks from the
Arithmetic Mean
can be interpreted differently by different
persons. The calculations may lead to Percentage of marks Deviations from A.M. (= 68.2)
different results. Further, if it does not 65 –3.2
use all values of the set it cannot be 59 –9.2
representative of the whole set of values. 75 + 6.8
Algebraic treatment of averages is useful 79 + 10.8
85 + 16.8
for further work. 45 –23.2
55 –13.2
3. Arithmetic Mean 70 + 1.8
77 + 8.8
Suppose ten students of your class 72 + 3.8
secured the following percentage of Total 0
marks in the Class X examination.
65 59 75 79 85 to the concept of ‘centre of gravity’ in
45 55 70 77 72 physics.

The arithmetic mean of marks is Example 1


obtained by adding up the marks and
(a) Verify that the sum of deviations from
dividing the total by the number of
any other value is not equal to zero.
students, i.e.
(b) What is the sum of deviations from
682 ‘zero’? (Deviations from zero are also
A.M. = = 68.2
10 called deviations from the origin.)
In general, suppose x1, x2, …, xn are n
Thus, the average percentage of
values of a certain variable x; then, their
marks is 68.2.
arithmetic mean is defined as
An interesting property of the
arithmetic mean is that x1 + x2 + ... + xn
the sum of deviations of values from n
which is symbolically expressed as x.
the arithmetic mean is zero.
Briefly, we describe this as
It is in this sense that the arithmetic
1
mean is the central value of the x = ∑xi
distribution. n
In the above example, the deviations where ∑ xi means the sum of n values.
of marks from the arithmetic mean are as
shown in Table 6.1. ∑ is a ‘Greek’ alphabet called ‘sigma’.
We note that the sum of positive In fact, this is capital sigma.
deviations from the arithmetic mean is
equal to the sum of negative deviations. Example 2
In other words, positive and negative
Verify that
deviations balance each other. In this
sense the arithmetic mean is analogous ∑( xi – x )= 0,
50 INTRODUCTION TO STATISTICS

i.e. the sum of deviations from the mathematical properties of the


arithmetic mean is zero. arithmetic mean, and
Let us write. (b) it can be calculated unambi-
guously from the given data.
∑ (xi – x ) = (x1 – x ) + (x2 – x ) + ... + (xn – x ) (iv) In calculating the arithmetic mean we
= (x1 + x2 + ... + xn) – n x use all values of the variable. This
means that even a single very large or
=0
very small value of the variable will
1 unduly affect the arithmetic mean.
n∑ i
because x = x
For example, the daily expenditure (in
rupees) of 5 households in a locality are
Example 3
25 28 32 27 33
The monthly expenditures, in rupees, of
5 households are The arithmetic mean of these values
is Rs 29.
1550 1715 1690 820 1150
Now, suppose a rich household
The arithmetic mean is 1385, i.e. the moves in this locality, whose daily
average monthly expenditure of these expenditure is Rs 125. If we recalculate
households is Rs 1385. the arithmetic mean, we get its value as
Note that the arithmetic mean may Rs 45. Since, the arithmetic mean has
not be equal to any one of the values of increased by more than 50 per cent one
the set. might think that the locality has become
richer. However, we observe that 5 out 6
4. Characteristics of Arithmetic Mean households are spending exactly the
same amount as before. The increase in
Let us now consider the special features
the value of the arithmetic mean is
of the arithmetic mean, which will help in
entirely because of the new arrival.
comparing it with other measures of
central tendency.
5. Median
(i) As stated above, the sum of
deviations of values from the Arithmetic mean is the central value of
arithmetic mean is zero. the distribution in the sense that positive
(ii) Suppose we are considering the and negative deviations from the
monthly incomes of some arithmetic mean balance each other. On
households. Then the arithmetic the other hand, median is the central
mean will give us that income which value of the distribution in the sense that
each household would get, if the total the number of values less than the
income was equally distributed median is equal to the number greater
among all households. than the median.
(iii) The arithmetic mean is a simple Suppose there are 17 students in
function of the values of the variable. your class. They secured the following
This has two advantages: percentage of marks in the Class X
(a) it is easy to derive the examination last year.
MEASURES OF CENTRAL TENDENCY 51

35 63 61 29 54 46 48 57 43 Example 4
60 39 40 51 68 38 56 53
The heights (in cm) of six students in your
If we arrange the marks in ascending class are
order of magnitude, we get
160, 157, 156, 161, 159, and 162.
29 35 38 39 40 43 46 48 51
53 54 56 57 60 61 63 68
What is the median height?
If we arrange heights in ascending
It follows that 8 students secured less order, we get
than 51 per cent marks and 8 secured
156 157 159 160 161 162
more than this percentage. Thus,
Median = 51 per cent. The two middle most values are the
1
Median is the central value in a sense 3 and 4th. The median = (159 + 160) =
rd
2
different from the arithmetic mean. In 159.5. Compare this with the arithmetic
case of the arithmetic mean it is mean = 159.2 (approx.).
the numerical magnitude of the
deviations that balances. But, for the 6. Characteristics of Median
median it is the number of values greater The main features of the median are as
than the median which balances against follows.
the number of values less than the (i) It is simple and easy to understand.
median. (ii) Like the arithmetic mean, median is
In general, if we have n values of x, rigidly defined.
they can be arranged in ascending order (iii) It is not affected by very large or very
as small values.
(iv) The median cannot be expressed as
x1 < x2 < … < xn. a simple function of the values of the
Suppose n is odd, then given set.

7. Mode
( )
Median = the n + 1 -th value.
2 Mode is defined as the value of the
However, if n is even, we have two variable which occurs most frequently.
middle values, viz., the The following is the frequency array
of ages of students in a cricket team in

(n2)-th and (n2 + 1)-th. your school.


TABLE 6.2
Frequency Array of Ages of Students
In this case,
Age Number of students
n -th value + n
Median =
()
2 2 ( )
+1 -th value 14
15
2
2
2 16 4
17 2
i.e. the arithmetic mean of the two middle 18 1
most values. Total 11
52 INTRODUCTION TO STATISTICS

There are 4 students of age 16. Since, expressed as a simple function of the
this is the most common age, with largest given values.
frequency, modal value is 16 years.
Sometimes, mode is more meaningful 9. Calculation of the Arithmetic Mean
average than the arithmetic mean or the for Ungrouped Data
median. Consider the following example.
Direct Method
Example 5 Let x 1, x 2 , …, x n be n values of some
variable x. The values have not been
A company which manufactures shoes for
grouped into class intervals. Then, by
adult men would like to concentrate on the
direct method, the arithmetic mean of x
production of shoes of only one size. What
is obtained as
should this size be?
1
n∑ i
The company conducted a survey of x= x;
2000 customers regarding their shoe size
and obtained the following distribution. i.e. we simply add up all values of x and
divide the sum by their number.
TABLE 6.3
Frequency Distribution of Shoe Size Assumed Mean Method
Shoe size Number of customers
Let A be the assumed mean. We obtain
5 55 the deviations of x from A, i.e.
6 164
7 360 d1 = x1 – A, d2 = x2 – A, ..., dn = xn – A.
8 528
9 410 Next, obtain the arithmetic mean of
10 260 deviations
11 137
12
Total
86
2000 d=
∑di
n
In this example, neither the and
arithmetic mean nor the median would be
an appropriate average. It would be better x = A + d.
to use mode. Since the largest frequency In other words, obtain the arithmetic
is 528 the mode is 8. It would be advisable mean of the deviations (d) and add A,
for the company to specialize in the which was earlier subtracted.
manufacture of shoes of size 8.
Step Deviation Method
8. Characteristics of Mode
Firstly, obtain deviations (d) from an
(i) Like the arithmetic mean and median, assumed mean.
mode too has an easy interpretation. Divide the deviations by a common
(ii) Mode is not affected by extreme factor, say h, to obtain
values because it depends only on the d1 d2 dn
largest frequency. d1' = , d2 ' = , ..., dn' =
(iii) Mode, like the median, cannot be h h h
MEASURES OF CENTRAL TENDENCY 53

Then gives
x = A + hd' x = 5550 + 1690 = 7240,
as before.
where
1 Step Deviation Method
n∑ i
d' = d' .
We divide the deviations (d) by a common
In other words, obtain the arithmetic factors, say 1000; and obtain
mean of d1', d2', …, dn'; then multiply it by
d1'=1, d2'=2, d3'=4, d4'= –1 and d5'= 2.45
h (because we had originally divided by h)
and add A (because we had earlier Now
subtracted A). ∑d'i 8.45
d' = = = 1.69;
5 5
Example 6 and
x = A + 1000 d'
Suppose the monthly incomes (x), in = 5550 + 1000 × 1.69
rupees, of 5 individuals, are given as = 7240
follows: as before.
6550, 7550, 9550, 4550 and 8000. We should note that taking deviations
Obtain their arithmetic mean. from an assumed mean, as well as the
step deviation method, help in reducing
Direct Method the burden of calculations a great deal.
They are particularly useful, if the
By direct method, we simply add all numbers are large.
incomes and divide by the number of
individuals. Thus, we obtain 10. Calculation of Arithmetic Mean for
36200 a Frequency Array
x= = 7240,
5 The variable x is discrete and takes values
i.e. the arithmetic mean of incomes of 5 x1, x2, …, xn with frequencies f1, f2, …, fn.
individuals is Rs 7240. (No class intervals are specified.)
Let us write the frequency array as
Calculation of the Arithmetic Mean by TABLE 6. 4
Assumed Mean Method Frequency Array

Let the assumed mean be A = 5550. Values of Frequency


x f
Obtain deviations (d) from the
x1 f1
assumed mean as x2 f2
d 1 = 1000, d 2 = 2000, d 3 = 4000, : ”
d4 = –1000 and d5 = 2450, xn fn

So that ∑ di = 8450 and d = 1690. Total N = ∑fi


Thus, x takes the value x1, f1 times
Then
x takes the value x2, f2 times
x=A+d x takes the value xn, fn times
54 INTRODUCTION TO STATISTICS

Therefore, by the Direct Method, the Example 7


arithmetic mean of x is
A survey of 100 households was carried
fi times f2 times fn times
out to obtain information on their family
{
{
{
x1+...+x1+x2+...+x2+...+xn+...+xn size. The results of the survey are
x=
N classified as a frequency array in the table
f1x1 + f2x2 + ...+ fnxn below.
=
N
∑fixi
TABLE 6.5
Frequency Array of Size of Households
=
N Size of the household Number of households
where N = ∑fi x f
1 5
Calculation of Arithmetic Mean by Assumed 2 15
Mean Method 3 25
4 35
Let A be the assumed mean. 5 10
Obtain the deviations of the values of 6 5
7 3
x from the assumed mean A: 8 2
d1 = x1 – A, …, dn = xn - A Total 100

Next, obtain
1 Calculate the arithmetic mean of size
d = ∑ fidi of households.
N
and Direct Method
x = A + d.
Let us arrange the calculations as in the
Calculations of Arithmetic Mean by Step following table.
Deviation Method TABLE 6.6
Calculation of Arithmetic Mean
Let h be the common factor. (By Direct Method)
Divide each deviation (d) by h, and
obtain x f f×x
d1 dn 1 5 5
d1' = , ..., dn' = 2 15 30
h h 3 25 75
Then 4 35 140
1 5 10 50
d' = N∑ fid'i 6 5 30
7 3 21
and 8 2 16
x = A + hd' Total 100 367
In words, we multiply d' by h
(because we had divided by h earlier), Therefore, the arithmetic mean is
and add A (because earlier we had 367
x= = 3.67
subtracted A). 100
MEASURES OF CENTRAL TENDENCY 55

Assumed Mean Method Direct Method


Let the assumed mean of x be 4. Let us arrange the calculations in the
The calculations are arranged as in following table.
the table below.
TABLE 6.9
TABLE 6.7
Calculation of Arithmetic Mean Calculation of Arithmetic Mean
(By Assumed Mean Method) (By Direct Method)

x d=x–4 f f×d Income No. of Individuals f×x


x f
1 -3 5 -15
2 -2 15 -30 11,500 629 7,233,500
3 -1 25 -25 12,000 1705 20,460,000
4 0 35 0
5 1 10 10 12,500 1800 22,500,000
6 2 5 10 13,000 725 9,425,000
7 3 3 9 13,500 420 5,670,000
8 4 2 8 14,000 127 1,778,000
Total 100 -70 + 37 = -33 14,500 85 1,232,500
15,000 9 135,000
Therefore,
Total 5500 68,434,000
33
d=– = – 0.33 and
100
Therefore, the arithmetic mean of
x = 4 + d = 4 – .33 = 3.67 incomes
as before.
x= ∑fixi =
68,434,000
≈ 12,442.55
In this example there is no extra N 5,500
advantage in using the step deviation
method, because the values of x are
Assumed Mean Method
already small and deviations from
assumed mean are convenient. Let the assumed mean be 12,500. We
Let us consider another example. arrange calculations in the table below.

Example 8 TABLE 6.10


Calculation of Arithmetic Mean
Let the frequency array of monthly (By Assumed Mean Method)
incomes of individuals be as follows. (Assumed Mean = 12,500)

TABLE 6.8 Income Deviations f f×d


Frequency Array of Incomes x d = x – 12500
Income (in rupees) Number of individuals 11,500 -1000 629 -629,000
11,500 629 12,000 -500 1705 -852,500
12,000 1705 12,500 0 1800 0
12,500 1800 13,000 500 725 362,500
13,000 725
13,500 420 13,500 1000 420 420,000
14,000 127 14,000 1500 127 190,500
14,500 85 14,500 2000 85 170,000
15,000 9 15,000 2500 9 22,500
Total 5500 Total 5,500 - 316,000
56 INTRODUCTION TO STATISTICS

Therefore, Example 9
1 316000
d = ∑ fidi= – ≈ – 57.45 The frequency distribution of the monthly
N 5500
expenditure of 100 rural households is
and
given below.
x = A + d = 12442.55.
TABLE 6.12
Step Deviation Method Frequency Distribution of Monthly
Expenditure of 100 Households
Let us now calculate the arithmetic mean Monthly expenditure in rupees Number of
of incomes by step deviation method. The (class intervals) households
calculations are shown in the following
75 – 125 10
table. 125 – 175 22
TABLE 6.11 175 – 225 38
Calculation of Arithmetic Mean 225 – 275 16
(By Step Deviation Method) 275 – 325 8
(Assumed Mean = 12500 and Common Factor 500) 325 – 375 4
375 – 425 2
Income Deviations d f f × d' Total 100
x d = x – 12500 d' =
500
11500 -1000 -2 629 -1258 Calculate the arithmetic mean of
12000 -500 -1 1705 -1705
12500 0 0 1800 0 monthly expenditures by
13000 500 1 725 725 (a) direct method
13500 1000 2 420 840 (b) assumed mean method, and
14000 1500 3 127 381
14500 2000 4 85 340 (c) step deviation method.
15000 2500 5 9 45

Total 5500 -632


(a) Direct Method

Therefore, Let us arrange the calculations in the


1 632 following table.
d' = ∑ fidi'= – ≈ – 0.1149
N 5500
TABLE 6.13
and
Calculation of Arithmetic Mean
x = A + hd' (By Direct Method)
= 12500 – 500 × 0.1149
Class interval Mid-value of f f×x
= 12442.55
class interval
We should note that the step x
deviation method simplifies large (1) (2) (3) (4)
calculations considerably.
75 – 125 100 10 1000
125 – 175 150 22 3300
11. Calculation of Arithmetic Mean – 175 – 225 200 38 7600
for Grouped Data 225 – 275 250 16 4000
275 – 325 300 8 2400
We assume that x is a continuous 325 – 375 350 4 1400
variable and the values of x are grouped 375 – 425 400 2 800
into equal class intervals. Total 100 20500
Let us illustrate calculations by using
the following example. Mid-values are obtained as
MEASURES OF CENTRAL TENDENCY 57

75+125 125+175 175+225 TABLE 6.15


=100, =150, =200, Calculation of Arithmetic Mean
2 2 2
(By Step Deviation Method)
(Assumed Mean = 200 and
and so on. ∑fixi = 20500 Common Factor = 50)
The arithmetic mean =
∑ fio 100 Class Mid- d = x-200
d' =
d f f × d'
interval value 50
= 205. x
(1) (2) (3) (4) (5) (6)
(b) Assumed Mean Method
Suppose the assumed mean = 200. Then 75 – 125 100 -100 -2 10 -20
the calculations of arithmetic mean are 125 – 175 150 -50 -1 22 -22
as shown in the following table. 175 – 225 200 0 0 38 0
225 – 275 250 50 1 16 16
TABLE 6.14 275 – 325 300 100 2 8 16
Calculation of Arithmetic Mean 325 – 375 350 150 3 4 12
(By Assumed Mean Method)
375 – 425 400 200 4 2 8
(Assumed Mean = 200)
Total 100 10
Class Mid-value Deviations f f×d
interval x d = x – 200 Therefore,
(1) (2) (3) (4) (5)

75 – 125 100 -100 10 -1000 d' =


∑f1d'i =
10
=
1
= 0.1
125 – 175 150 -50 22 -1100 N 100 10
175 – 225 200 0 38 0 and
225 – 275 250 50 16 800
x = A + hd' = 200 + 50 × 0.1 = 205.
275 – 325 300 100 8 800
325 – 375 350 150 4 600
375 – 425 400 200 2 400 where A = 200 and the common factor
Total 100 500 h = 50.

Therefore, 12. Calculation of Median

d=
∑fidi = 500 = 5 Ungrouped Data
N 100 Given a set of values of the variable, we
arrange them in ascending (or,
and
descending) order of magnitude. Median
x = A + d = 200 + 5 = 205. is the middle most value.
If the number of values is odd, say 101,
(c) Step Deviation Method
median is the 101 + 1 = 51-st value. In
Looking at the Column (2) of the above 2
general, if N = 2m+1, the median is the
Table 6.14, we see that the common
(m + 1)-th value. Otherwise, if N = 2m is
factor is 50. In fact, this is the width of even, there are two middle most values,
class intervals. viz. the m-th and (m+1)-th values. Median
The calculations are as shown in the is the arithmetic mean of the two.
following table. We have considered examples before.
58 INTRODUCTION TO STATISTICS

Example 10 TABLE 6.17


Cumulative Frequencies
The following data give the number of
Class Frequency Less than cumulative
passengers carried by a bus in ten trips. Interval frequencies
(1) (2) (3)
22 26 14 30 18
11 35 41 12 32 75 – 125 15 15=15
125 – 175 22 37=15+22
Verify that the arithmetic mean of 175 – 225 38 75=15+22+38
number of passengers carried is 24.1. 225 – 275 16 91=15+22+38+16
275 – 325 8 99=15+22+38+16+8
In order to find the median number, 325 – 375 5 104=15+22+38+16+8+5
we arrange the numbers in ascending 375 – 425 3 107=15+22+38+16+8+5+3
order as Total 107

11 12 14 18 22 26 30 32 35 41 The values of x are already arranged


There are two middle values, viz., the in ascending order in the frequency table.
5th and the 6th. Median is the arithmetic We observe that 15 values are less than
mean of the two, i.e. 125; 15+22=37 values are less than 175;
15+22+38=75 values are less than 225,
22 + 26 48 and so on. These are called ‘less than’
median = = = 24.
2 2 cumulative frequencies.
Grouped Data Since x is a continuous variable, the
N 107
Example 11 -th value (i.e. = 53.5)-th is the
2 2
Suppose x is a continuous variable, and the middle most value, i.e. the median.
the values have been grouped into class Since 37 values are less than 175 and
N
intervals. We have the frequency 75 values are less than 225, the -th
2
distribution shown in the table below. value (i.e. 53.5 -th value) will lie in the class
(175-225).
TABLE 6.16 Thus, (175 -225) is the median class.
Frequency Distribution
We assume that all values are
Class interval Frequency uniformly distributed in any class.
x f Now, use the formula given below to
75 – 125 15 interpolate the median, i.e. the 53.5-th
125 – 175 22 value: N
175 – 225 38 – c.f.
225 – 275 16 2
275 – 325 8
Median = L + ×h
f
325 – 375 5 where
375 – 425 3 L = Lower Limit of the median class (i.e.
Total 107 = 175)
N = total frequency (i.e.107)
Calculate the median value of x. c.f.=number of values less than the lower
In order to calculate the median, first limit L (i.e. 37)
of all, calculate the cumulative frequencies f = frequency in the median class (i.e. 38)
as shown in the following table. h = width of the median class (i.e. 50)
MEASURES OF CENTRAL TENDENCY 59

Therefore, substituting values in the Height (in feet) No. of students


above formula
3.0 – 3.5 12
53.5 – 37
Median = 175 + × 50 = 196.7 3.5 – 4.0 37
38 4.0-4.5 79
In order to compare this value with 4.5-5.0 152
the arithmetic mean, see that the 5.0-5.5 65
arithmetic mean in this case is 203.27. 5.5-6.0 7
Total 352
13. Calculation of Mode
Since 152 is the largest frequency,
Mode is that value of x for which the the modal class is (4.5 – 5.0). Thus, L =
frequency is maximum. For example, if x
4.5, f0 = 152, f1 = 79, f2 = 65 and h = 0.5,
takes the values 70, 69, 69, 75,75, 75, 60
so that
and 80, the mode (or, the modal value) is
clearly 75. 152 – 79
If the values of x are grouped into Mode = 4.5 + × 0.5
2 × 152 – 79 – 65
classes (such that they are uniformly
distributed within any class) and we have = 4.73 ft. (approx.).
a frequency distribution,
(i) identify the class which has the 14. Relative Position of the Arithmetic
largest frequency, and Mean, Median and Mode
(ii) calculate the mode as
Suppose we use the following notations
f0 – f1
Arithmetic Mean = Me
Mode = L + ×h Median = Mi
2 f0 – f1 – f2
Mode = Mo
where
where e, i and o are the suffixes. Then the
L = Lower limit of the modal class relative magnitude of the three are either
f0 = Largest frequency of the following :
f1 = frequency in the class preceding the (i) Me > Mi > Mo,
modal class (ii) Me < Mi < Mo
f2 = frequency in the class next to the suffixes occurring in alphabetical order.
modal class The median is always between the
h = width of the modal class. arithmetic mean and the mode.
Observe that if f1 = f2,
1 15. Partition Values — Quartiles
Mode = L + h
2
i.e. mode is equal to the middle value of Median divides the total set of values of
the modal class. the variable in two equal parts. Similarly,
the quartiles divide the total set of values
Example 12 into four equal parts.
There are three quartiles Q1, Q2 and
Data relating to the height of 352 school Q3.
students are given in the following Q1 is called the lower quartile.
frequency distribution. The number of values less than Q1
Calculate the modal height. is one-fourth and more than Q 1 is
60 INTRODUCTION TO STATISTICS

three-fourths of the total number of values N = total frequency


in the set. c.f.= number of values less than L
Q3 is called the upper quartile f = frequency in the class in which Q1
Three-fourths of the total number of lies,
values in the set are less than Q3 and one- h = width of the class in which Q1 lies.
fourth are greater than Q3. In our example
Q 2 is the median or the second N
quartile. L = 25, = 75, c.f. = 23, f = 64 and h = 25.
4
Thus, the number of values between
Q 1 and Q 2 is one-fourth of the total Therefore, substituting in the above
number of values; and so is the number formula Q1 = 45.3125.
between Q2 and Q3. We follow a similar procedure to
The number of values between Q1 and interpolate Q3.
Q3 is one-half of the total number of 3
values in the entire set. Q3 is the()
4
N -th value
While calculating quartiles we 3
assume that the values of the variable are Since N = 300, N = 225.
4
uniformly distributed in any class. We have to locate the 225-th value.
From the frequency distribution, we
Example 13
see that 202 values are less than 75 and
Calculate the quartiles Q1 and Q3 for the 253 values are less than 100. Therefore,
following frequency distribution. the 225-th value must lie in the class 75-
100.
Class interval Frequency Cum frequency
Q3 is calculated by using a similar
0-25 23 23 formula as above for Q1.
25-50 64 87
3
50-75 115 202 N – c.f.
75-100 51 253 4
100-125 30 283
Q3 = L + ×h
f
125-150 17 300
Total 300 where
L = lower limit of the class in which Q3
The total frequency = N = 300, lies,
N N = total frequency
therefore, =75.
4 c.f. =number of values less than L
We find that 23 values are less than f = frequency in the class in which Q3
25 and 87 are less than 50. Therefore, the lies
75 th value (Q 1) must lie in the class
interval 25-50. h = size of the class in which Q3 lies.
Q1 is calculated by using the formula In the present case, Q3 lies in the class
interval 75-100.
N
– c.f. We have
4
Q1 = L + ×h 3
where f L=75, 4 N=225, c.f.=202, f=51 and h=25.
L = lower limit of the class in which Q1 Therefore, substituting in the above
lies, formula Q3 = 86.2745.
MEASURES OF CENTRAL TENDENCY 61

EXERCISES

1. Compare the arithmetic mean, median and mode as measures of central tendency.
Describe situations where one is more suitable than the others.
2. The arithmetic mean is described as the centre of gravity of the distribution of
values of the variable. Explain.
3. Show that the sum of deviations of the values of the variable from their arithmetic
mean is equal to zero.
4. If x1, …, xn are n values of the variable x and A is any arbitrary value, so that
∑(xi – A) is the sum of deviations from A; show that ∑(xi – A) = n(x – A) where
1
x = ∑xi is the arithmetic mean of x. Hence ∑(xi–A) = 0 if and only if, A = x.
n
5. ‘Arithmetic mean is affected by very large and very small values, but median and
mode are not affected by them. Explain?
6. Daily expenditures on vegetables (in rupees) of 20 households, in a certain locality,
are given below.
25.00 26.50 30.25 28.00 23.00
31.40 34.00 33.00 30.50 27.20
28.00 35.00 38.60 34.00 22.50
24.00 23.70 28.00 29.00 32.20
Show that the value of the median lies between the arithmetic mean and the mode.
7. Calculate the arithmetic mean, median and mode for the following frequency
distribution.
Class Interval Frequency
1-3 3
3-5 10
5-7 16
7-9 13
9-11 8
11-13 3
13-15 1
Total 54
Examine their relative positions. Does the median lie between the arithmetic
mean and mode?
8. Out of 50 questionnaires distributed among 50 manufacturing firms only 35
replies were received. In the column inquiring about ‘capacity utilization’ the
figures read.
(In percentages)
54.2 96.8 74.2 57.8 99.7 84 81.2
94.3 93.7 61.6 100.0 110.1 68 71
95.2 98.3 84 91.9 84 94.1 73
84 52.6 49.2 45 62.6 75.1 84
69.2 90.7 85.4 87.3 66.1 73.9 89.3
62 INTRODUCTION TO STATISTICS

(a) Calculate the arithmetic mean, median and mode, and examine their relative
position.
(b) Group the data into appropriate classes, and again calculate the arithmetic
mean, median and mode from grouped data. How do your results differ from
those obtained in (a)? Explain.
9. In a certain examination there were 100 candidates of whom 21 failed, 6 secured
distinction, 43 were placed in the third division and 18 in the second division. It
is known that at least 75 per cent marks are required for distinction, at least 40
per cent for passing, at least 50 per cent for second division and at least 60 per
cent for first division.
Calculate the median of the distribution of marks.
10. Calculate the upper and lower quartiles for the following frequency distribution.
Class interval Frequency
13-25 6
25-37 11
37-49 23
49-61 7
61-73 3
Total 50
11. Calculate the median for the distribution in the Question No.10. Also, derive the
median geometrically, by drawing the two ogives and locating the point of
intersection.
12. The daily expenditure, in rupees, of 50 households is given as follows.
Daily expenditure (Rs) Number of households
100-150 3
150-200 9
200-300 21
300-500 10
500-1000 5
Above 1000 2
Total 50

[In this example the last class is open. Therefore, we cannot determine the mid-
value of that class. The arithmetic mean cannot be calculated for the given data.
However, we may set an upper limit on some basis and then obtain the arithmetic
mean. But, this must be regarded as an approximation. Calculation of median,
mode and quartiles do not pose any difficulty.]
(i) Calculate the arithmetic mean, median and mode. Compare the values.
(ii) Also, compute the upper and lower quartiles.

You might also like