I am trying to analyze an experiment with the design in the title:
2(between) x 2(within) x 2(within--but only manipulated in one group)
My initial hunch was to run two separate 2 (between) x 2 (within) ANOVA.
However, this approach feels suspect to me.
Can I analyize this effectively in one model?
To give some context, this is a memory experiment with 16 stimuli, which were word-picture pairs.
The between-subject variable is type of memorization technique: passive reading or practice testing (e.g., word - ?).
One within-subject variable is type of final memory test prompt (e.g., word - ? or ?-picture).
The other within-subject variable only pertains to the practice-testing group. This is whether practice-test prompt matches the final test prompt (e.g., word-?/word-? OR ?-picture/?-picture) or does not match (word-?/?-picture OR ?-picture/word-?).
Thus, it's 2(learning technique: reading, pratice testing) x 2 (final-test prompt: word - ?, ?-picture) x 2 (practice-test/final-test match: match, no match).
Any help would be appreciated!