Fisher's Exact Test
Fisher's exact test is a test for independence between the row and column variables of a contingency table. When the data consist of two categorical variables, a contingency table can be constructed reflecting the number of occurrences of each factor combination. Fisher's exact test assesses whether the value of one factor is independent of the value of the other. For example, this might be used to test whether political party affiliation is independent of gender.
Certain types of homogeneity (for example, homogeneity of proportions in a k by 2 table) are equivalent to the independence hypothesis. Hence, this test may also be of interest in such cases.
As this is an exact test, the total number of counts in the cross-classification table cannot be greater than 200. In such cases, the chi-square test of independence is preferable.
To perform Fisher's exact test
Choose Statistics Compare Samples
Counts and Proportions
Fisher's Exact. The dialog shown below appears.
The Fisher's Exact Test has the following options:
Data
Data Set
Specify a data set.
Variable 1
Specify the column from the data set which offers the first classification or grouping variable. This must be a factor or category.
Variable 2
Specify the column from the data set which offers the second classification or grouping variable. This must be a factor or category.
Data Set is a Contingency Table
Select if the data set specified is a contingency table.
Results
Enter the name for the object in which to save the results of the analysis.
Print Results
Select this to print out the results of the analysis in the designated output window.
Related S-PLUS language Functions
fisher.test, print.htest, menuFisher