0
$\begingroup$

I've built a multivariate linear mixed effect model with 3 dependent variables and a repeated measures factor. My data follows a hierarchical structure where participants are nested within groups.

my model is as follows:

model <- lmer(value ~ variable:Time -1 + (0 + variable | Group/Participant), data = Fixed_Data

I'm attempting to calculate the overall r-squared for this model, and have been following this post.

As such, i've used this code:

r.squaredGLMM(model)

I'm a little iffy about this though, as I'm receiving an r-square of 90+%

To compare, I've calculated this multivariate analysis as three individual univariate analyses:

model1 <- lmer(DV1 ~ Time + (1|Group/Participant), data = Data) ...

When calculating the r-squared for these models, I'm getting r-squares of under 10% each.

Is the discrepancy between these r-square values normal? Is calculating an r-square for a multivariate linear mixed effect model possible?

Thanks!

$\endgroup$
  • $\begingroup$ Did you try different packages? Eg does performance::r2() give similar or identical results? $\endgroup$ – Daniel Jun 3 at 20:08

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.