These are studies that have mulitple factors at one time.

Example

• 24 subjects
• they use either Apple Iphone or Samsung Galaxy keyboard on phone
• they use it when they are either sitting, standing or walking
• measure: WPM, error rate (% of each phrase entered)
• entries are 20 phrases selected by us
• They’ll write the 20 phrases in the three postures
• We are also interested in seeing how the different postures affect the performance. This is called interaction

This is a 2x3 mixed factorial design. This means we have 2 factors, one with 2 levels (keyboards) and the other has 3 levels (postures).

It is a mixed factorial design because keyboard is a between-subject factor, and posture is a within-subject factor. (example-analysis-of-an-ab-test)

The responses are WPM and the error rate %.

In an experiment of this kind, we would record every entry, with one entry being the performance of a single phrase.

We can fully counterbalance the postures factor since it is just 6 combinations.

References

• anova-assumptions
• one-way-anova
• interaction-effects