I am measuring the X, Y, Z variables of i subjects in k states, and for each subject and state i am taking j measurements. What is the proper way to test for differences between the k states?

It seems likes a one way repeated measure ANOVA problem, however how should I deal with the fact that in each subject and within each state i take j measurements of X, Y, Z? Should I take the mean of the Xj, Yj, Zj within each subject and state and with this mean do a one way repeated measures ANOVA?

Many thanks?


I think you want a mixed model (aka multilevel model, hierarchical linear model, etc). Instead of modeling $Y = XB + e$ with $e \sim \mathcal{N}(0, \sigma)$ these model $Y = XB + Z\gamma + e$

I have to run, but this should give you places to look

| cite | improve this answer | |

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.