The Friedman test is a nonparametric statistical test used to detect differences across multiple related groups. It is often used as an alternative to repeated measures ANOVA when the assumptions of normality or sphericity are not met. The test ranks the data within each block (such as each participant) and then analyzes the differences in the ranks across the groups.
Each block must contain one observation for every group, meaning the test requires a complete set of repeated measures from each subject or unit. The number of blocks (such as participants) should be reasonably large—typically at least five blocks are recommended—to provide enough power for the test to detect differences.
For example, in a study where the same group of students rates their satisfaction with three different teaching methods, the Friedman test could be used to determine whether there are significant differences in satisfaction without assuming normal distributions of the ratings. Each student (block) would rate all three teaching methods (groups), and the rankings would be compared across students.
« Back to Glossary Index