I have a data set which includes sex, age, and 5 polygenic scores as independent variables, with 16 dependent variables. I have constructed univariate linear mixed effects regression models and corrected for multiple comparisons using the FDR method.

I've also constructed a multivariate model including all the the predictors. Some of the polygenic scores have a significant effect on dependent variable X in the multivariate but not univariate models.

I have a few questions about this:

A) Do I need to correct for multiple comparisons if I have many multivariate linear mixed effects models?

B) It was suggested to me that I use a likelihood ratio test to compare the model with just age and sex to the model with all 5 predictors added at the same time. What unique information would this give me practically?

C) When I ran the likelihood ratio test, I got an insignificant p value for a model which did have a predictor showing a significant effect on the dependent variable. Why this discrepancy?

D) How would I report all of these results? Should I list the effect size and p value from the full models and also the results of the likelihood ratio test?


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.