LESSON 5

Measures of Variability

INTRODUCTION:
In the previous lessons, you learned how to compute mean, median and
mode. These measures of centrality focus only on giving information of what
score could best represent the entire set of data. However, if you want to
determine the spread of scores, the measure of variability can address that
query. Thus, in this lesson you will be learning the different kinds of measures of
variability or sometimes called as measures of dispersion.

OBJECTIVES:
:At the end of this lesson, you should be able to:
1. Compute and interpret the various measures of variability such as the
range, variance, standard deviation and coefficient of variation for a set of
ungrouped data.
2. Compute and interpret the various measures of variability such as the
range, variance, standard deviation and coefficient of variation for a set of
grouped data.

Range
Range is the crudest measure of dispersion. It is the difference between
the highest and the lowest scores in the data set. This means that range
considers only two scores, thus making it the most unstable measure of
dispersion. For ungrouped data range is
R  HL
where:
R – range
H – highest score
L – lowest score

For example, find the range of the two sets of data below.
1. 3, 4, 5, 5, 6, 8, 10, 15
Solution:
R = 15 – 3 = 12
 2. 89, 70, 37, 40, 30, 95, 77, 25, 53, 36
Solution:
R = 95 – 25 = 70

Range of Ungrouped Data Using MS Excel

The MS Excel has no built-in formula to compute for the range of data.
However, since range is just the difference between the highest and the lowest
scores, we can find the range of the data set using the max and min functions of
MS Excel. Thus, the steps will be as follow:
1. Open MS Excel.
2. Enter the data.
3. Insert “=max(data set)-min(data set)” in the formula bar.
4. Press “Enter”.

Insert the formula.

Enter the data.

After pressing the “Enter” key, the result will be 12, which is the same result as
in manual computation. MS Excel will be most helpful if you have hundreds of
entries.

Range of Grouped Data

The range of grouped data will be the difference between the highest
upper limit and the lowest lower limit among the class intervals. In formula,

R  HU  L L
where:
R – range
HU – highest upper limit
LL – lowest lower limit

For example, considering the frequency distribution table given above.

 Class Intervals f
90 - 94 5
85 - 89 7
80 - 84 10
75 - 79 15
70 - 74 10
65 - 69 5
60 - 64 3

Solution:
The highest upper limit among the class intervals is 94 and the lowest
lower limit is 60. Therefore, the range will be

R  94  60  34

Interquartile Range and Quartile Deviation

Other measures of variability that use quartiles are the interquartile range
(IQR) and Quartile Deviation (QD). Both IQR and QD use Q1 and Q3. Here are
there formulas:

IQR  Q 3  Q 1
Q 3  Q1
QD 
2

Example: Find the IQR and QD of ungrouped data: 2, 3, 3, 4, 5, 5, 6, 6, 7, 8.

Solution:
n  1 10  1 11
Find Q1: The position of Q1 is at    2.75
4 4 4
Therefore: Q1  3  .75(3  3)  3
3(n  1) 3(10  1) 33
Find Q3: The position of Q3 is at    8.25
4 4 4
Therefore: Q 3  6  .25(7 - 6)  6.25

Since we computed already the values of Q1 and Q3, we can now plug in the
values in our formulas.

IQR  Q 3  Q1  8.25  2.75  5.5

Q 3  Q1 8.25  2.75
QD    2.75
2 2
Find the IQR and QD of grouped data below:

Class Intervals f
36 - 40 4
31 - 35 6
26 - 30 10
21 - 25 15
16 - 20 20
11 - 15 16
6 - 10 4

Solution:
The very first thing we have to do is to construct a cumulative frequency
column, identify the i and N.

Class Intervals f Cf
36 - 40 4 75
31 - 35 6 71
26 - 30 10 65
21 - 25 15 55
16 - 20 20 40
11 - 15 16 20
6 - 10 4 4
i=5 N=75

N 74
For Q1: Solve   18.75 and locate in the Cf column.
4 4
Class Intervals f Cf
36 - 40 4 75
31 - 35 6 71
26 - 30 10 65
21 - 25 15 55
16 - 20 20 40
LB=10.5 11 - 15 16 20
6 - 10 4 4
i=5 N=75
 N 
 4  Cfb 
Q1  LB   i
 f Q1 
 
 75 
 4  4 18.75 - 4 
Q1  10.5   (5)  10.5   (5)  15.11
 16   16 
 

3(N) 3(74)
For Q3: Solve   56.25 and locate in the Cf column.
4 4
Class Intervals f Cf
36 - 40 4 75
31 - 35 6 71
LB=25.5 26 - 30 10 65
21 - 25 15 55
16 - 20 20 40
11 - 15 16 20
6 - 10 4 4
i=5 N=75

 3N 
 4  Cfb 
Q 3  LB   i
 f Q 3 
 
 3(75) 
 4  55  156.25 - 55 
Q 3  25.5   (5)  25.5   (5)  26.125
 10   10 
 

Therefore:
IQR  Q 3  Q1  26.125 - 15.11  11.015
Q 3  Q1 26.125  15.11
QD    5.51
2 2

Mean Absolute Deviation

Mean absolute deviation (MAD) or simply average deviation is the average
distance of all scores from the mean. Since, the average deviation of all the
mean scores from the mean is always zero, MAD uses the absolute value to get
rid of useless zero deviation.
Mean Absolute Deviation of Ungrouped Data
In general, the mean absolute deviation is given
n

xx
MAD  i1
for sample and
n

x m 
MAD  i1
for population.
N

where:  – summation symbol

 – absolute value symbol
x – individual score
x - sample mean
Xm - midpoint
- population mean
n – number of scores in sample
N – number of scores in population

For example, we are going to find the MAD of the following scores: 3, 4, 5, 5, 6,
8, 10, 15.

Solution:
It is easier to compute MAD if we are going to construct a table.
Step 1. Construct the table for easy computation.
x
3
4
5
5
6
8
10
15

Step 2. Compute the mean.

3  4  5  5  6  8  10  15 56
x  7
8 8
Step 3. Construct x - x column.
x xx
3 -4
4 -3
 5 -2
5 -2
6 -1
8 1
10 3
15 8

Step 4. Construct x  x column.

x xx xx
3 -4 4
4 -3 3
5 -2 2
5 -2 2
6 -1 1
8 1 1
10 3 3
15 8 8
Step 5. Get the sum of x  x column.
x xx xx
3 -4 4
4 -3 3
5 -2 2
5 -2 2
6 -1 1
8 1 1
10 3 3
15 8 8
xx  24

Step 6. Substitute in the formula.

MAD 
xx 
24
3
n 8

Mean Absolute Deviation of Ungrouped Data Using MS Excel

In computing the mean absolute deviation using MS Excel, these are the
following steps:

Step 1. Open MS Excel Program.

Step 2. Type in your data.
Step 3. In formula bar, type “=AVEDEV(data range)”.
 Type the formula here!

Type in your data.

Step 4 click “Enter”.

As you can see, the result is 3. This is equal to the result of the manual
computation.

Mean Absolute Deviation of Grouped Data

To compute the mean absolute deviation of grouped data, the formula will
be

f i xi  x
MAD  i1

n
where:

 - summation symbol
fi – frequency of ith class
xi – midpoint of ith class
x - mean of distribution
n – number of cases

Let us take the frequency distribution table given above and compute the MAD:

Class Intervals f
90 - 94 5
85 - 89 7
80 - 84 10
75 - 79 15
70 - 74 10
65 - 69 5
60 - 64 3
Solution:

Step 1. Compute the mean. In the previous lesson, the mean of the given data is
77.91.

Step 2. Construct x m  x column.

Class Intervals f xm xm  x
90 - 94 5 92 14.09
85 - 89 7 87 9.09
80 - 84 10 82 4.09
75 - 79 15 77 -0.91
70 - 74 10 72 -5.91
65 - 69 5 67 -10.91
60 - 64 3 62 -15.91

Step 3. Construct x m  x column.

Class Intervals f xm xm  x xm  x
90 - 94 5 92 14.09 14.09
85 - 89 7 87 9.09 9.09
80 - 84 10 82 4.09 4.09
75 - 79 15 77 -0.91 0.91
70 - 74 10 72 -5.91 5.91
65 - 69 5 67 -10.91 10.91
60 - 64 3 62 -15.91 15.91

Step 4. Construct f x m  x column.

Class Intervals f xm xm  x xm  x f xm  x
90 - 94 5 92 14.09 14.09 70.45
85 - 89 7 87 9.09 9.09 63.63
80 - 84 10 82 4.09 4.09 40.9
75 - 79 15 77 -0.91 0.91 13.65
70 - 74 10 72 -5.91 5.91 59.1
65 - 69 5 67 -10.91 10.91 54.55
60 - 64 3 62 -15.91 15.91 47.73

Step 5. Get the sum of f x m  x .

Class Intervals f xm xm  x xm  x f xm  x
90 - 94 5 92 14.09 14.09 70.45
85 - 89 7 87 9.09 9.09 63.63
80 - 84 10 82 4.09 4.09 40.9
 75 - 79 15 77 -0.91 0.91 13.65
70 - 74 10 72 -5.91 5.91 59.1
65 - 69 5 67 -10.91 10.91 54.55
60 - 64 3 62 -15.91 15.91 47.73
 f x m  x  350.01
Step 6. Substitute in the formula.

MAD 
f x m x

350.01
 6.36
n 55

Variance and Standard Deviation

Other measures of variability are the variance and standard deviation.
Unlike mean absolute deviation which uses the absolute value symbol, variance
and standard deviation square the difference of scores and the mean.
In fact, variance is defined as the average squared deviation from the mean
while standard deviation is the square root of variance.

Variance of Ungrouped Data

The formula for population and sample variance are as follow:

Population variance, σ 2

 (x  μ) 2

and sample variance, s 2


 (x  x) 2

N n -1
where: σ 2 - is the population variance
s2 – sample variance
x – individual score
 - population mean
x - sample mean
N – number of scores (population)
n - number of scores (sample).

Standard Deviation of Ungrouped Data

For standard deviation, since it the square root of variance, the formula for
population and sample standard deviation will be:

Population standard deviation, σ 

 (x  μ) 2

and sample standard deviation,

N

s
 (x  x) 2

.
n -1

where: σ - is the population standard deviation

 s – sample standard deviation
x – individual score
 - population mean
x - sample mean
N – number of scores (population)
n - number of scores (sample)

Since the variance and standard deviation are the measures of variability or
spread, they are interpreted as the lower the value the more clustered the scores are
and the higher the value the more spread the scores are.
Suppose a sample of 10 children has the following ages: 4, 5, 5, 6, 6, 7, 8, 9, 9,
11. Let us try to compute the variance and the standard deviation of the ages of these
10 children.

The solution will be:

Step 1. Compute the mean.

4  5  5  6  6  7  8  9  9  11 70
x  7
10 10

Step 2. Construct x  x column.

x xx
4 -3
5 -2
5 -2
6 -1
6 -1
7 0
8 1
9 2
9 2
11 4
Step 3. Construct (x  x) 2 column.

x xx (x  x) 2
4 -3 9
5 -2 4
5 -2 4
6 -1 1
6 -1 1
7 0 0
 8 1 1
9 2 4
9 2 4
11 4 16
 (x  x) 2
 44

Step 4. Substitute in the formula:

For variance: s 2

 (x  x) 2

n -1

s2 
 (x  x) 2 
44

44
 4.89
n -1 10  1 9
For standard deviation:

s
 (x  x) 2


44

44
 4.89  2.21
n -1 10  1 9
Alternate Formula for Variance and Standard Deviation of Ungrouped
Data

This is the alternative formula for the variance and standard deviation of
ungrouped data:
n x 2   x 
2

For variance, s 2
and
n(n  1)
n x 2   x 
2

For standard deviation, s 

n(n  1)

Let us take the same problem as given above:

Solution:

Step 1. Construct the x2 column.

x x2
4 16
5 25
5 25
6 36
6 36
7 49
8 64
9 81
9 81
 11 121
Step 2. Get the sum of x column and the sum of x2 column.
x x2
4 16
5 25
5 25
6 36
6 36
7 49
8 64
9 81
9 81
11 121
 x  70 x 2
 534
Step 3 Substitute in the formula.

For variance:
n x   x 
2
2 10(534)  (70) 2 5340  4900 440
s      4.89
n(n  1) 10(10  1) 10(9) 90
For standard deviation:
s  4.89  2.21
These values were the same as the result of the previous computation using
different formula.

Variance and Standard Deviation of Ungrouped Data Using MS Excel

Using MS Excel is capable of computing variance and standard deviation of
ungrouped data. Let us try to compute the variance and standard deviation of
the data given above using MS Excel.

To compute for sample variance, these are the steps:

Step 1. Open MS Excel and type in the data.

Step 2. Enter “=VAR.S(data range)” in the formula bar.

You will see that the value is 4.89.

For sample standard deviation, the formula should be “=STDEV.S(data range)”.

After pressing the “Enter” button, MS Excel will give you 2.21 as an answer.

Variance and Standard Deviation of Grouped Data

The formula for variance and standard deviation for grouped data is given
as follow:
 For variance, s 2

 (x m -x) 2
and
n 1

For standard deviation, s 

 (x m -x ) 2
n 1
Let us compute the variance and standard deviation of the data given above:
Class Intervals f xm xm  x
90 - 94 5 92 14.09
85 - 89 7 87 9.09
80 - 84 10 82 4.09
75 - 79 15 77 -0.91
70 - 74 10 72 -5.91
65 - 69 5 67 -10.91
60 - 64 3 62 -15.91

Solution:

Step 1. Compute the mean. The mean is equal to 77.91 as computed previously.

Step 2. Construct x m  x column.

Class Intervals f xm xm  x
90 - 94 5 92 14.09
85 - 89 7 87 9.09
80 - 84 10 82 4.09
75 - 79 15 77 -0.91
70 - 74 10 72 -5.91
65 - 69 5 67 -10.91
60 - 64 3 62 -15.91

2
Step 3. Get the square of x m  x and put it on (x m  x) column.
Class Intervals f xm x m  x (x m  x) 2
90 - 94 5 92 14.09 198.5281
85 - 89 7 87 9.09 82.6281
80 - 84 10 82 4.09 16.7281
75 - 79 15 77 -0.91 0.8281
70 - 74 10 72 -5.91 34.9281
65 - 69 5 67 -10.91 119.0281
 60 - 64 3 62 -15.91 253.1281

Step 4. Set the compute  f(x m  x) 2 .

Class Intervals f xm xm  x (x m  x) 2 f(x m  x) 2
90 - 94 5 92 14.09 198.5281 992.641
85 - 89 7 87 9.09 82.6281 578.397
80 - 84 10 82 4.09 16.7281 167.281
75 - 79 15 77 -0.91 0.8281 12.4215
70 - 74 10 72 -5.91 34.9281 349.281
65 - 69 5 67 -10.91 119.0281 595.141
60 - 64 3 62 -15.91 253.1281 759.384
 f(x m  x)  3454.55
2

Step 5. Substitute in the formula.

For variance, s 2 
 (x m -x) 2

3454.55
 63.97 and for standard deviation,
n 1 55 - 1
s  63.97  7.998 .

Alternate Formula for Variance and Standard Deviation of Grouped

Data
The alternate formula for variance and standard deviation of grouped data
are as follow:

n fx m   fx m 
2 2

For variance, s  2
and
n(n  1)
n fx m   fx m 
2 2

For standard deviation, s 

n(n  1)

Let us take the same example as above and compare the results.

2
Class Intervals f xm fx m fx m
90 - 94 5 92 460 42320
85 - 89 7 87 609 52983
80 - 84 10 82 820 67240
75 - 79 15 77 1155 88935
70 - 74 10 72 720 51840
65 - 69 5 67 335 22445
60 - 64 3 62 186 11532
  fx  fx
2
N  55 m  4285 m  337295

Solution:

n fx m   fx m 
2 2
2 55(337295)  4285 2 18551225  18361225
s   
n(n  1) 55(55  1) 55(54)
190000
  63.97
2970

s  63.97  7.998

Coefficient of Variation
The standard deviation is not reliable measure to compare two data sets in
terms of spread when the two data sets are of different units or have the same
units but widely dissimilar mean in the field. In this case, the coefficient of
variation is developed to answer this kind of problem. The formula for coefficient
of variation is given below:
s
CV 
x
where:
CV – coefficient of variation
S – standard deviation
x - mean

The data 4, 5, 5, 6, 6, 7, 8, 9, 9, 11 which was previously given has a x  7 and

s  2.21. Therefore,
s 2.21
CV    0.3157
x 7