I am currently analysing data of a cohort study where we try to model change in a dependent variable (say, academic grades) over three time-points based on a number of continuous independent variables.

The change in grades occurs due to a single intervention between time point 1 and 2 (say, tutoring). Hence, grades increase between TP1 to TP2 and then remain stable to TP3. Independent variables are related to the intervention: weekly hours of tutoring; relationship with the tutor; satisfaction / engagement with the tutoring program.

Due to a substantial amount of missing data at outcomes and for predictors variables, I am planning to use a linear MIXED model, entering grades as DV and the predictors as covariates in SPSS.

Now where I am uncertain: In order to interpret the effect of the predictor variables on change (of grades), is it sufficient to include the main effects of time and predictors (as covariates) in the fixed part, OR is it necessary to look at the interactions between time and the respective predictors?

My understanding is that the main effects can only be interpreted as influencing the outcome across time points,so if there was a main effect of satisfaction on grades, this could also just mean that people who are satisfied with their tutoring already have high grades at TP1, is that correct?

Thanks very much in advance


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.