1.3.5.9.

F-Test for Equality of Two Variances

Page 1 of 3

1. Exploratory Data Analysis

1.3. EDA Techniques
1.3.5. Quantitative Techniques

1.3.5.9. F-Test for Equality of Two Variances

Purpose:
Test if
variances
from two
populations
are equal

An F-test (Snedecor and Cochran, 1983) is used to test if the

variances of two populations are equal. This test can be a two-tailed
test or a one-tailed test. The two-tailed version tests against the
alternative that the variances are not equal. The one-tailed version
only tests in one direction, that is the variance from the first
population is either greater than or less than (but not both) the second
population variance. The choice is determined by the problem. For
example, if we are testing a new process, we may only be interested
in knowing if the new process is less variable than the old process.

Definition

The F hypothesis test is defined as:

12 = 22
H0:
Ha:
2 < 2 for a lower one-tailed test
1
12
12

Test
Statistic:

>

2
22
22

for an upper one-tailed test

for a two-tailed test

F = s21 /s22

where s21 and s22 and are the sample variances. The
more this ratio deviates from 1, the stronger the
evidence for unequal population variances.
Significance
Level:
Critical
The hypothesis that the two variances are equal is
Region:
rejected if

F > F,N1 1,N2 1

F < F1,N1 1,N2 1
F < F1/2,N1 1,N2 1

for an upper one-tailed test

for a lower one-tailed test
for a two-tailed test

or

F > F/2,N1 1,N2 1

http://www.itl.nist.gov/div898/handbook/eda/section3/eda359.htm

7/27/2015

1.3.5.9. F-Test for Equality of Two Variances

Page 2 of 3

where F, N1-1, N2-1 is the critical value of the F

distribution with N1-1 and N2-1 degrees of freedom and
a significance level of .
In the above formulas for the critical regions, the
Handbook follows the convention that F is the upper
critical value from the F distribution and F1- is the
lower critical value from the F distribution. Note that
this is the opposite of the designation used by some
texts and software programs.
F Test
Example

The following F-test was generated for the AUTO83B.DAT data set.
The data set contains 480 ceramic strength measurements for two
batches of material. The summary statistics for each batch are shown
below.
BATCH 1:
NUMBER OF OBSERVATIONS
MEAN
STANDARD DEVIATION

=
=
=

240
688.9987
65.54909

BATCH 2:
NUMBER OF OBSERVATIONS
MEAN
STANDARD DEVIATION

=
=
=

240
611.1559
61.85425

We are testing the null hypothesis that the variances for the two
batches are equal.
H0:
Ha:

12 = 22
12 22

Test statistic: F = 1.123037

Numerator degrees of freedom: N1 - 1 = 239
Denominator degrees of freedom: N2 - 1 = 239
Significance level: = 0.05
Critical values: F(1-/2,N1-1,N2-1) = 0.7756
F(/2,N1-1,N2-1) = 1.2894
Rejection region: Reject H0 if F < 0.7756 or F > 1.2894

The F test indicates that there is not enough evidence to reject the
null hypothesis that the two batch variancess are equal at the 0.05
significance level.
Questions

The F-test can be used to answer the following questions:

1. Do two samples come from populations with equal variancess?
2. Does a new process, treatment, or test reduce the variability of
the current process?

Related
Techniques

Quantile-Quantile Plot
Bihistogram

http://www.itl.nist.gov/div898/handbook/eda/section3/eda359.htm

7/27/2015