Analyses

Fisher’s exact test

Exact test of association for a 2×2 table.

Edit on GitHub

When to use it

Use Fisher’s exact test on a 2×2 table when expected counts are small. It asks the same question as chi-square: are the two binary variables associated?

Assumptions

Both columns have exactly two levels. The test conditions on the row and column totals.

Running it in Tensr

Analyze → Descriptive Statistics → Crosstabs → Exact. In chat: “Fisher’s exact test of group by outcome.”
Column A and column B must each have two levels.
The same test runs if you turn on Fisher’s exact inside chi-square.

Options

Prop

Type

Reading the output

A 2×2 of exposure and case status. The case rate was built only a little higher in the exposed row, on 90 people.

exposedNoYes
No2815
Yes2324

Fisher's exact test for exposed × case, n = 90. The primary result is not significant (p = .140). This is not large enough to treat the comparison this page is about as a reliable association. The result is non-significant: the data are still compatible with no effect. Metrics: Odds ratio = 1.948; 95% CI = [0.834, 4.552]; p (two-tailed) = .140; p (one-tailed) = .140.

Reporting (APA 7)

Fisher's exact test for exposed × case, n = 90. This result is not significant (p = .140). Report the estimate with that p, and do not describe the pattern as a reliable effect.

Coming from SPSS

Crosstabs → Exact → Fisher’s exact test. SPSS also prints a one-sided p. Tensr stores both p_value_two_tailed and p_value_one_tailed.