# Fisher's exact test

## Description

Fisher's exact test is a significance test for a 2x2 table. This test evaluates all distribution probabilities for a 2 x 2 table and produces an exact probability for a given set of observed frequencies. The null hypothesis is that the row and column variables are unrelated, or that there is no difference in the respective proportions.

## Required input

- Variable 1 and 2: select categorical or qualitative variables. These variables may either contain character or numeric codes. These codes are used to break-up the data into a two-way classification table.
- Optionally select a filter to include a subset of cases.

## Graph

Select one of the following:

- Simple column chart (one classification factor). The chart contains a single bar for each category. The height of the bars is the number of cases in the category.
- Clustered column (two classification factors). Like simple column chart, but containing a group of bars for each category in the first classification category. The height of the bars is the number of cases in the category.
- Stacked column (two classification factors). Bar segments are stacked on top of one another. There is one bar stack for each category in the first classification factor. Segments within each stack represent the contribution of categories in the second classification factor.
- 100% Stacked column (two classification factors). Bar segments are stacked on top of one another, the total equals 100%. There is one bar stack for each category in the first classification factor. Segments within each stack represent the relative contribution of categories in the second classification factor.

## Results

The results window for Fisher's exact test displays:

**Frequency table**: the program displays the 2x2 classification table.

When you select the option **Show all percentages** in the results window, all percentages relative to row, column and grand totals, are shown in the table.

**P-value**: when the (two-sided) P-value (the probability of obtaining the observed result or a more extreme result) is less than the conventional 0.05, the conclusion is that there is a significant relationship between the two classification factors, or that there is a significance difference in the respective proportions.

## See also

## Link

Go to Fisher's exact test.