we have an experiment with 3 within-subject factors: SOA(4 levels), Congruency (2 levels) and Visibility Rating (4 levels). RT is our dependent variable. Because we receive Visibility Rating score from participants for each trial, we have a number of missing values for some conditions/participants, for example, a participant can never use Visibility rating 1, so we miss all SOA*Congruency conditions for visibility rating 1. (Yes, we still want to use Visibility rating in the model, though technically it is not a regular independent variable). Therefor, we want to used Mixed models for repeated measures design. We did found a variety of codes for similar designs (but only for 2 within- and 1 between-subject factors), we also noticed that people use different syntax in proc Mixed. Could you, please, help us and suggest an appropriate code for our particular case in SAS? We will greatly appreciate any feedback. Dina.

| cite | improve this question | | | | |
  • $\begingroup$ Welcome to CV. This site is not intended to be a source for questions about specific software or coding issues. Your question should be reworded to reflect the more general issues raised. You might consider browsing Judith Singer's paper, Using Proc Mixed to Fit Multilevel Models, which is available in an ungated version for some possible direction. Finally, it would help potential respondents in formulating their answers if you were not to assume that everyone knows what "SOA" or "RT" refers to. In other words, some explanation rather than acronyms and buzz words would help your case. $\endgroup$ – Mike Hunter Mar 27 '16 at 13:26

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.