MEASURES OF CENTRAL TENDENCY AND DISPERSION
In addition to graphical summaries, the primary features of a data set can be summarized through
numerical indices. One of such indices is what we generally refer to as measures of central
tendency.
Definition: Measures of central tendency are those of the location of the middle or the centre of a
distribution. They are statistics that attempt to describe typical scores that reflect how the data is
similar. They enable one to be more objective when drawing conclusions.
The 3 most commonly used measures of central tendency are:
(i) The mean
(ii) The mode
(iii) The median
The mean is defined as the sum of scores, measurements or values divided by the number of
times they occur in a given distribution or data set. It is obtained by adding up all the scores and
dividing the sum by the number of times they occur in the given data set.
Example: Calculate the mean of the following scores by ten students in a statistics test:
55, 62, 72, 58, 81, 73, 66, 77, 80, 65
.=55+62+72+58+81+73+66+77+80+65
10
= 689/10
= 68.9
The mode is defined as the most frequent score in a distribution. It is the score which occurs
with the highest frequency in a given distribution.
Example: Find the mode of the following data set:
55, 62, 72, 55, 81, 66, 77, 55, 80, 65.
The mode is: 55.
The median is the score found in the middle of data that is arranged in either ascending or
descending order. It is the value that divides the data set in a given distribution into two equal
halves.
1
Example: Calculate the median score of the following data set:
113, 133, 95, 112, 94.
By arranging the data set in either ascending or descending order, the median score is: 112.
NB: In case there are 2 middle scores in a distribution arranged in either ascending or descending
order, the median value is obtained by calculating the average of the two middle scores.
Example: Calculate the median score of the following data set:
105, 102, 101, 92, 91, 80
By arranging the scores in either ascending or descending order, the middle values are: 101 and
92.
= 101+92
= 193/2
= 96.5
CALCULATING THE MEAN, MODE AND MEDIAN OF A GROUPED FRQUENCY
DISTRIBUTION
The mean:
In calculating the mean of a grouped frequency distribution, we use the mid-points of the class
interval and its associated frequency. We apply the following formula:
¯x = ∑fx/∑f
Where ¯x = mean
f = the frequency of the class
x = the mid-point of the particular class
Example:
2
Calculate the mean of the data set given below:
Class Interval Frequency Mid-point fx
(f) (x)
90-99 40 94.5 3780
80-89 90 84.5 7605
70-79 127 74.5 9461.5
60-69 148 64.5 9546
50-59 117 54.5 6376.5
40-49 87 44.5 3871.5
30-39 45 34.5 1552.5
20-29 32 24.5 784
10-19 14 14.5 203
∑f = 700 ∑fx= 43180
The mean = ∑fx/∑f
= 43180/700
= 61.6857
The mode:
The mode is calculated using the following formula:
M = L + i (d1/d1+d2)
Where M = mode
L = the lower real limit of the modal class
i = the class interview or width
d1 = the difference between the frequency of the modal class and the pre-modal class
d2 = the difference between the frequency of the modal class and post-modal class.
3
Example: Use the above grouped frequency distribution to calculate the mode.
M=?
L = 59.5
i = 10
d1 = (148-117) = 31
d2 = (148-127) = 21
M = L + i (d1/d1+d2) (USE THE BODMASS RULE)
= 59.5 + 10 (31/31+21)
= 59.5 + 10 (31/52)
= 59.5 + 10 (0.5961)
=59.5 + (10x0.5961)
=59.5 + 5.96
. = 65.46
NB: In case a distribution has two or more classes with modal frequencies, you can either
calculate only one of the modes or all the corresponding modes ( bi-modal or multi-modal)
distributions.
The median:
The median for a grouped frequency distribution is calculated using the cumulative frequency
and apply the following formula:
Md = L + i (N/2-cfb)/fw
where Md = Median
L = the lower real limit of the median class
i= the class width or interval
N= ∑f = the number of scores
4
Cfb = cumulative frequency before the median class
fw= the frequency of the median class.
Example: Use the above grouped frequency distribution to calculate the median value.
L=49.5
i=10
N=700
Cfb=178
fw=117
Md = L+ i(N/2-cfb)/fw
= 49.5 + 10 (700/2-178)/117
= 49.5 + 10 (350-178)/117
= 49.5 + 10 (172/117)
= 49.5 + 10 x 1.47
= 49.5 + 14.7
= 64.2
5
MEASURES OF DISPERSION:
Measures of dispersion—also known as measures of variability—refer to the extent to which
scores in a given distribution deviate from the central tendency. They are used to establish the
deviations of the scores from the centre.
The most common measures of dispersion or variability are:
(i) The range
(ii) The variance
(iii) The standard deviation
The range:
In simple terms, the range is the difference the largest score or measurement and the smallest or
lowest score or measurement in a given distribution.
Example: Calculate the range of the following data set:
160, 130, 100, 70, 40, 30, 80, 105, 120, 114
The range = 160-40
= 120.
The range of a grouped frequency distribution can be calculated in two ways:
(i) By considering the difference between the mid-points of the largest class and the mid-
points of the smallest class.
In the above frequency distribution, the range = 94.5 -14.5 = 80.
(ii) By considering the difference between the upper real limit of the largest class and the
upper real limit of the smallest class.
In the above frequency distribution, the range = 99.5 -19.5 = 80.
The variance:
This is a measure that shows us how spread out a particular distribution is in a given data set. It
is computed using the following formula:
δ2 = ∑fx2/∑f-(∑fx/∑f)2
6
Example: Calculate the variance of the following grouped frequency distribution showing the
ages of Social Statistics students at Makerere University:
Ages of students Frequency
(f)
17-21 12
22-26 15
27-31 7
32-36 4
37-41 2
Clue: Remember to calculate the values of x, x2, fx and fx2 before applying the above formula.
The Standard deviation:
This statistic tells us how tightly or loosely the scores or measurements are clustered around the
mean in a given data set. It is kind of the “mean of the mean” and often can help you find the
story behind the data. It is the square root of the sum of the squared deviations of the scores from
the mean divided by the number of scores.
The standard deviation is calculated using the following formula:
δ or s = √ ∑fx2/∑f-(∑fx/fx)2
OR
δ or s = √ N∑fx2-(∑fx)2/N(N-1).
Exercise: Use the above grouped frequency distribution to calculate the standard deviation.
Clue: As above.
7
8
