Track 10 · Comparing Groups and Categorical Data · Lesson 7 of 11
Repeated measures foundations
Analyse a within-subject design with one-way repeated-measures ANOVA, explain why removing subject-to-subject variation increases power, and describe the sphericity assumption and its corrections.
A courier company measures the same 60 routes before and after a routing change; a design team times the same six testers on three checkout designs. Each subject is measured under every condition, so differences between subjects — long routes, slow testers — can be removed from the comparison instead of drowning it.
In a between-subjects design each subject contributes to one condition. In a within-subject (repeated-measures) design each subject is measured under every condition. Observations from the same subject are correlated, so the one-way ANOVA of the previous lessons — which assumes independent observations — is not valid. But the correlation is also an opportunity: each subject acts as their own control.
Included with a Formula Quest membership
It has 7 graded exercises with feedback, hints and worked solutions. After it you will be able to:
- Recognise within-subject designs and why their observations are not independent
- Partition variation into conditions, subjects and error, and compute the repeated-measures F
- Explain sphericity and apply a Greenhouse–Geisser or lower-bound correction to the df
- Use the paired t-test (k = 2) or the Friedman test as related analyses