Fisher's exact test is used to analyze contingency tables that contain exactly two rows and two columns, and that contain less than 50 cases. Unlike the chi-square test, Fisher's exact probability is not plagued by inaccuracies due to small N's. Therefore, it should be used instead of chi-square for 2x2 contingency tables where the total N is less than 50. The probability will be similar to the chi-square test with Yate's correction.
Example
Here are the results of a recent "favor or oppose" public opinion poll broken down by gender. What is the exact probability that the difference between the observed and expected frequencies occurred by chance?
| Male | Female |
|---|
| Favor | 2 | 3 |
|---|
| Oppose | 1 | 33 |
|---|
Fisher's exact probability = .037
You could make a statement like this: Fisher's exact test was calculated on a 2x2 table examining the relationship between opinion and gender. The probability was significant at the .05 alpha level, p=.037. Therefore, we conclude that there was a statistically significant relationship between respondents' opinions and their gender.