Yes, I Think that a Chi-squared goodness of fit suitable for you.

Hi! Given the binary outcome of interest being recall or no recall, you can include musical genre (sounds like there are four different genres) to determine the odds of recall using logistic regression in SPSS (genre is the predictor and recall is the dependent response). You can certainly do this with a Pearson chi-square test also based on a 4 x 2 table (genre in the 4 rows and yes/no recall as the 2 columns) to assess the association or effect of genre. The advantage of a logistic regression approach is that you can include other variables in the analysis, which you may also want to examine as possible factors associated with the outcome.

Hello Luke,

The answer depends on how you quantify recall. If it is a score, perhaps how many passages of a given genre were correctly recalled, then you'd likely be better off treating recall as a (pseudo) continuous score, and stick with the repeated measures anova. In addition to genre as an IV/factor in the study, you could separate recall scores by: (a) early/late series recall scores vs. (b) middle series recall scores, thus adding a second repeated measures factor.

One downside to taking the information from continuous scores and reducing them to a dichotomous variable (e.g, recall was better in early/late series vs. recall was not better in early/late series) is that you end up treating any score difference as "real," no matter how small.

Good luck with your work.