When data are not normally distributed, the Kruskal-Wallis test in SPSS provides a robust alternative to one-way ANOVA. In this exercise, you will examine reaction times under three conditions and learn how to check the assumptions, correctly apply the Kruskal-Wallis test, and interpret the results – including an appropriate post-hoc test. Ideal for reviewing and consolidating your knowledge of SPSS or PSPP.
Task
A researcher examines participants’ reaction times under three different conditions. The data are not normally distributed.
| Condition | Reaction times (ms) |
|---|---|
| 1 | 250, 270, 260, 255, 265 |
| 2 | 300, 310, 305, 295, 315 |
| 3 | 280, 275, 285, 290, 295 |
Task:
- Check the assumptions for conducting a one-way ANOVA in PSPP or SPSS.
- Since the assumptions are not met, conduct a Kruskal-Wallis test instead.
- If the result is significant, conduct an appropriate post-hoc test and interpret the results.
If you still can’t solve the task, review the material on the one-way ANOVA in SPSS and the Kruskal-Wallis test again.
Solution with PSPP or SPSS
Solution video
Screenshot of the solution
