I am planning to run a mixed-effects model using the lme4 package. I will have two fixed effects (group and condition) and three random effects (item, subject, and list). In considering whether to include random slopes along with my random intercepts, it seems to me that of the six possible combinations of fixed effects and random effects, five could make sense as slopes and intercepts, where the fixed effect corresponds to the slope and the random effect corresponds to the intercept. For example, I could include a random intercept for subject and a random slope for condition to go with the random intercept for subject. All the examples I've seen have been of this nature, where the intercept corresponds to a random effect and the slope corresponds to a fixed effect. My question is whether it's possible / makes sense to include an intercept and slope where both correspond to a random effect, for example, in my case, a random intercept for subject and a random slope for item. Conceptually this seems to make sense to me, since the dependent variable could vary by subject, and additionally, any effect of item on the dependent variable may vary differently by subject, but since item is not a fixed effect, I'm wondering if what I'm considering doesn't make sense statistically?

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.