The F-Test Two-Sample for Variances tool tests the null hypothesis that two samples come from two independent populations having the equal variances.

In the example below, two sets of observations have been recorded.  In the first sample, students were given a test before lunch and their scores were recorded.  In the second sample, students were give a test after lunch and their scores recorded. 

F-Test Two-Sample for Variances Example Dataset

To run the test: 

  1. On the XLMiner Analysis ToolPak pane, click F-Test Two-Sample for Variances
  2. Enter A1:A16 for Variable 1.
  3. Enter B1:B16 for Variable 2. 
  4. Keep "Labels" selected since our two columns are labeled with "Before Lunch" and "After Lunch"
  5. Leave the default value for Alpha.
  6. Select cell D1 for Output Range.    
  7. Click OK. 

F-Test Two-Sample for Variances Pane

The results are below.

F-Test Two-Sample for Variances Results

If the calculated F value in cell E8 is less than the F critical value in cell E10, then the null hypothesis (the variances are equal) is accepted. Since the calculated F is larger than the F Critical value, we will reject the null hypothesis, the variances for the two samples are not equal.