CHAPTER 4
VARIABILITY
Educational Statistics
EDU 5950
By Chan Yoke Bee (GS37395)
VARIABILITY
Provides a quantitative measure of the
differences between scores in a distribution
and describes the degree to which the
scores are spread out or clustered together.
Graph population distribution
mean
mean
PURPOSE FOR MEASURING VARIABILITY
To obtain an objective measure of how the scores
are spread out in a distribution.
A good measure of variability serves two purpose:
A) variability describes the distribution
-- closer or spread out
~~ variability === distance
B) how well an individual / groups scores represent
the entire distribution.
** provide information about how much error
expected if using sample to represent a
population
MEASURE OF VARIABILITY
A) the
B)
range
standard deviation
C) variance
RANGE
Is the distance covered by the scores in a
distribution, form the smallest scores to largest
score.
Range
= Xmax - Xmin
=
When the scores are all whole number, the range
= Xmax - Xmin + 1
RANGE
Scores are measurement of a continuous variables,
range = upper real limit for Xmax lower real limit for Xmin
example: scores have the values from 1 to 5,
the range = 5.5 0.5 = 5
Scores are measurement of a discrete variables,
(when the scores are all whole number)
range = Xmax - Xmin + 1
example: the number of children in a family, data
produces
RANGE
ABC
3
2.5
1.5
ABC
0.5
0
0
RANGE
The problem with using the range as a
measure of variability:
It is completely determined by the two extreme
values and ignores the other scores in the
distribution
Thus, a distribution with one unusually large
(or small) score will have large range even
that is other scores are clustered close
together
Range often does not give an accurate
description, so the range is considered to be
a crude and unreliable measure of
STANDARD DEVIATION
Use the mean of the distribution as a
reference point and measures variability by
considering the distance between each score
and the mean
-- provide standard or average distance from
the mean
-- describe whether the scores are clustered
closely or widely scattered
STANDARD DEVIATION
Step 1
find the deviation (distance from the mean)
Deviation score
= distance from the mean
=X-
(X = score, = mean)
Example
a) = 50,
i) score is 53
ii) score is 45
deviation score = 53 50
deviation score = 45 50
=3
= -5
Deviation score can be sign ( + or -)
+ is the score located above the mean
- Is the score located below the mean
STANDARD DEVIATION
Step 2 find the mean of the deviation
scores
Add up the deviation scores and divide by N
Example X N = 4 scores,X- X = 12 = 12/4 = 3
8
+5
-2
-3
Deviation scores = (X
- )
STANDARD DEVIATION
Step 3
find the mean square of deviation
(variance)
Variance = average squared distance from
the mean
STANDARD DEVIATION
Step 4 Calculate the square root of the
variance
standard deviation =
Only apply in numerical scores interval or
ratio scale.
STANDARD DEVIATION
Example
Population = 1, 9, 5, 8, 7
5
Score, X
1
9
5
8
7
N = 5 scores
Deviation (X - )
Squared deviation, (X - )2
-5
25____________
3
9_____________
-1
1_____________
2
4_____________
1
____1______________
Sum of the squared deviation = 40
= 40
Variance (mean of the squared deviation) = 40 /5
=8
Standard deviation =
=
= 2.83
= 30 /
=6
Figure 4.3
frequency
=6
3
1
5
2
1
1
10
Estimate the standard deviation can help
you avoid errors in calculation
FORMULAS FOR
POPULATION
VARIANCE AND STANDARD DEVIATION
A) Sum of squared deviation scores (SS)
a)
b)
Definitional formula, SS =
or
Computational formula, SS =
Small groups of scores and the mean is whole number,
use definitional formula
When the mean is not a whole number, use
computational formula
FORMULAS FOR POPULATION VARIANCE AND STANDARD
DEVIATION
B) Variance =
C) Standard deviation =
Population variance,
Population standard deviation, =
EXAMPLE CALCULATION
Population A: 6,4,2,0,9,3
Find the sum of the squared deviation, SS
b)
Calculate the variance and standard deviation
Solution:
Score, X
X-
(X - ) 2
a)
Mean, = 24/6
=4
6
4
2
0
9
3
SS= (X - ) 2 = 50
variance = SS / N
= 50 / 6
= 8.3
Standard deviation = ( 50/6)
= 2.887
2
0
-2
-4
5
-1
Total,
4
0
4
16
25
1
50
STANDARD DEVIATION AND VARIANCE FOR
SAMPLES
SS is the same for a sample as it is for a population
Notation mean is M; number of scores is n
Steps almost are same:
A) deviation = X M
B) squared deviation =
C) Definitional formula, SS =
D) Computational formula, SS =
STANDARD DEVIATION AND VARIANCE FOR
SAMPLES
Sample variance, s2 =
Sample
standard deviation, s =
EXAMPLE 4.5
Population B: 1,6,4,3,8,7,6
Calculate the sample variance and standard deviation
TOTAL
X2
36
16
64
49
36
35
211
Solution:
Mean, M = 35/7 = 5
n= 7
SS =
= 211 175
= 36
Sample variance =
standard deviation =
=6
2.45
THE USES OF SAMPLE VARIABILITY
Sample variation = estimated population
variance
Sample standard deviation = estimated
population standard deviation
Degree of freedom, df determine the number of
scores in the sample that are independent and
free to vary
df = n 1
Without some correction, sample variability
consistently underestimates the population
variability. Dividing by a smaller number (n-1)
increases the value of the sample variance and
makes it an unbiased estimate of the population
variance
A sample statistic is unbiased if the average
value of the statistic is equal to the
population parameter. (The average value of
the statistics is obtained from all the possible
samples for a specific sample size, n)
A sample statistics is biased if the average
value of the statistics either underestimates
or overestimates the corresponding
population parameter.
TRANSFORMATIONS OF SCALE
1.
Adding a constant value to each scores
does not change the standard deviation
2.
Multiplying each score by a constant causes
the standard deviation to multiplied by the
same constant
THANK YOU..
