Can anyone guide me towards the proper analysis for a mixed design where the DV is categorical? I have a 2x2x2 design where two of the variables are within-subjects and one is between-subjects (so each participant rated four scenarios). The outcome is a 2-level categorical DV (e.g., yes/no). I'm not sure how to properly handle the nonindependence of the data when analyzing and would welcome suggestions.
Asked
Active
Viewed 185 times
1
-
You could look into a mixed model version of logistic regression. Relevant posts: http://stats.stackexchange.com/questions/81057/mixed-effect-logistic-regression-in-r-choosing-random-effects http://stats.stackexchange.com/questions/55877/interpretation-of-probabilities-from-a-mixed-model-logistic-regression http://stats.stackexchange.com/questions/226359/choosing-one-mixed-effects-model-in-a-logistic-regression http://stats.stackexchange.com/questions/105664/measuring-goodness-of-fit-for-mixed-logistic-regression-model-inconsistent-res – kjetil b halvorsen Sep 26 '16 at 14:08