Choose the test's alpha value. The alpha value represents the statistician's acceptable false positive rate -- that is, the rate at which the test rejects the statistician's hypothesis even though the hypothesis is true. For most applications, an alpha level of 5 percent is suitable, although some scientists use alpha values ranging between 1 and 10 percent. When in doubt, use 5 percent.
Determine or estimate the degrees of freedom for the first sample. Recall that an F-test needs two samples. Each sample has its own sample size (number of subjects). If you know the sample size of the first sample, you can determine the degrees of freedom. If you have not yet taken the sample, estimate the likely sample size (i.e., guess how many subjects you will be likely to include in your study from the first population). The degrees of freedom are the sample size minus one. Call this "df1."
Determine or estimate the degrees of freedom for the second sample. Do this in the same way that you did for the first sample. Call the degrees of freedom for this sample "df2."
Find the F-table corresponding to your value of alpha. Most introductory statistical texts include several F-tables. Each F-table corresponds to a different alpha level. Usually, statistical textbooks at least include F-tables for the alpha values: 1 percent, 5 percent and 10 percent. Find the table for your alpha value.
Find the critical F-value in the F-table. Go down the rows until you find your first degree of freedom. Go across the columns until you find your second degree of freedom. The intersection of that row and column is a cell containing a number that represents the critical F for your study. This critical F is the criterion for rejection in region F. If an F-statistic is greater than this F, rejection occurs.