08 November 2024 0 8K Report

Dear Researchers,

I am working on comparing two mixed models from the same dataset to determine if reductions in one variable over time are significantly correlated with reductions in another variable over time.

Specifically, I am studying whether a depression treatment not only reduces symptom severity but also lowers the concentration of an inflammatory marker, and I would like to assess whether these reductions are significantly correlated. For that purpose, I performed two different mixed models: one with depression severity as dependent variable and the other one with the concentration of the inflammatory marker as dependent variable. Time variable is included as independent variable.

Since I am unable to perform a multivariate mixed model with two dependent variables, I have researched alternatives and found that comparing slopes from separate models may help evaluate whether changes in these variables (symptom severity and inflammation) over time are correlated.

Therefore, I have a couple of questions:

  • Could you advise on how to compare slopes from mixed models in this case to determine if changes in the two variables over time are significantly correlated?
  • Are there other methods or analyses within mixed models that could help assess the correlation of changes in two variables over time, as I am describing it?
  • I am currently using Jamovi for the analysis, but I am open to recommendations for other software that may be suitable for this purpose.

    Thank you very much for your help!

    Best regards, Bruno

    More Bruno Pedraz's questions See All
    Similar questions and discussions