I'm running a lmer with a continuous outcome TestY measured in each visit a participant makes to the study i.e. a longitudinal repeated measures analysis. The outcome is a tablet-based test and the value used is z-scored.

I'm interested in the role of a specific categorical predictor Group and it's interaction with time.

Overall, TestY declines with time. However, it is known that this sample of study participants actually have an increase - due to a learning effect - between first and second time they take the test. Thus, it would be important to adjust for this known effect, which seems to be possible by adding a fixed offset term.

Let's say the known effect is that between 1st and 2nd visit we have a +0.33 between these visits. But that's it, no adjustments for future visits would be of interest (follow-up goes like to mean 6 visits).

My lmer looks like this:

lmer(TestY ~ Group + Group:Time_bl + covariates + (1 + Time_bl | ptID))

Question: Does anyone has some light to shed on how could I possibly account for this .33 increase between first and second visit?

It seems I'd probably have to add VisitNumber to the model, but I can't imagine how to code this offset term only between first and second visit.


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.