0
$\begingroup$

I have 6 outcome variables (each with data at multiple time points) and four predictor variables. I am aware I can use separate multiple regression analysis for each outcome variable for each time point.
I would like to understand what benefit multivariate analysis would bring? (I am struggling because of the mixed usage of multivariate when it should be multivariable!) Does multivariate analysis look at the relation between the outcome variables as well? I have a small sample size so unsure if multivariate analysis would be appropriate.

$\endgroup$

1 Answer 1

1
$\begingroup$

If the statistical model you choose provides sufficient accuracy, a multivariate model provides more insight and statistical power. An example of insight is obtaining estimates of correlations between outcomes. For Bayesian modeling you can compute posterior probabilities of compound statements involving multiple outcome tendencies. Power comes from combining information. For accuracy consider Bayesian models.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.