I'm running an ordinal mixed model on data collected from when my participants ranked 16 photos of people of who they think are most similar to them. Depending on the age of the participant, they would see different photos. My response variable is similarity ranking (1-4) and my predictors (fixed factors) are sex (M/F), age group (4-6, 7-9 and 10-12) and image category (1-4). I know that subject ID is a random factor however would image be as well since not all participants saw all images (48), but only a subset (16) based on their age?

  • $\begingroup$ In brief, yes...if you are not going to do a more complicated IRT (item response theoretic) model, then you should use crossed-level multi-level modeling...as you suggest, treat subject and item as random values at the 2nd (higher) level. $\endgroup$ – Gregg H Jul 10 at 13:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.