Fisher's exact test is a statistical significance test used to determine if there's an association between two categorical variables. Unlike the chi-square test, it's particularly useful when you have small sample sizes, especially when expected cell frequencies are less than 5. This guide will walk you through performing Fisher's exact test using SPSS.
Understanding When to Use Fisher's Exact Test
Before diving into the SPSS procedure, let's clarify when Fisher's exact test is the appropriate choice. You should consider it when:
- You have two categorical variables: Both variables should have only two or more categories.
- Your data is presented in a contingency table: This table shows the frequencies of observations for each combination of categories.
- Your sample size is small: This is the crucial criterion. When expected cell counts in a chi-square test are low (generally less than 5), Fisher's exact test provides a more accurate result.
Step-by-Step Guide: Performing Fisher's Exact Test in SPSS
Here's a step-by-step guide using a hypothetical example: let's say we're investigating the relationship between gender (male/female) and preference for coffee (yes/no).
1. Inputting Your Data:
First, enter your data into SPSS. You'll need two variables: one for gender (let's call it "Gender") and one for coffee preference ("CoffeePreference"). Each participant will have one entry for each variable. For example:
| Gender | CoffeePreference |
|---|---|
| Male | Yes |
| Female | No |
| Male | Yes |
| Female | Yes |
| Male | No |
| Female | No |
| ... | ... |
2. Creating a Contingency Table:
SPSS doesn't directly perform Fisher's test on raw data; it requires a contingency table. To create one:
- Go to Analyze > Descriptive Statistics > Crosstabs.
- Move your "Gender" variable into the "Row(s)" box and your "CoffeePreference" variable into the "Column(s)" box.
- Click Statistics.
- Check the box for Chi-square. Crucially, also check the box for Fisher's exact test.
- Click Continue, then OK.
3. Interpreting the Output:
SPSS will produce a contingency table showing the observed frequencies and the results of both the Chi-square and Fisher's exact tests. Focus on the Fisher's exact test results:
-
Exact Sig. (2-sided): This is the p-value. This is the most important value. If the p-value is less than your chosen significance level (typically 0.05), you can reject the null hypothesis. The null hypothesis states there is no association between the two variables. A significant p-value indicates a statistically significant association between gender and coffee preference.
-
Exact Sig. (1-sided): This is relevant only if you have a specific directional hypothesis (e.g., you hypothesize that males are more likely to prefer coffee). Most often, the two-sided test is appropriate.
Important Considerations
- Assumptions: Fisher's exact test doesn't assume large sample sizes like the Chi-square test. However, it still assumes independence of observations.
- Small Sample Sizes: While it's designed for small samples, extremely small samples (e.g., only a few observations in some cells) can still lead to limitations in the test's power.
- Alternative Tests: If your data violates the assumptions or you have more than two categories in one or both variables, consider alternative tests such as the Chi-square test (with caution if cell counts are low) or Barnard's exact test.
By following these steps, you can effectively utilize SPSS to conduct Fisher's exact test and interpret its results to determine the relationship between two categorical variables, particularly when dealing with limited sample sizes. Remember to always carefully consider the context of your research and the limitations of any statistical test.
