What do Covariates do in a mixed ANCOVA?

firona

Hi,

first of all, thank you for creating JASP :)

I’m currently in need of controlling for a variable in a mixed ANCOVA and I’m a little unclear how JASP includes the covariate in the model. If I, for instance, were to include baseline scores as a covariate, would JASP control for it in such a way, that its effect would have been controlled for in the group × time interaction?

I currently have two sources with contradicting information: one, which claims that the covariate only affects the between subject factor(s) and nothing more and the other saying that the covariate is also controlled for with regard to the interaction.

My question is: What is your take on this? As I need to control for the effect of a baseline value in one analysis and age in another, can I make this statement for my interaction after including these covariates in JASP?

Thanks! ☺️

Cheers,

Firona

EJ

Hi Firona,

I am pretty sure JASP includes the covariate as you would in R. So your question would then be on a more conceptual level, I assume? Intuitively I've always thought of the covariates as follows: you include only the covariates, and this yields residuals; you then account for these residuals with your regular model (similar to partial correlation). I'll ask some other people from the team to check this intuition. Maybe that intuition is only appropriate under specific circumstances, as suggested here: https://www.tandfonline.com/doi/abs/10.1080/03610910903243687?journalCode=lssp20

Also, you say that you need to control for baseline scores in one analysis and for age in another: so you are not just running one analysis with both baseline scores *and* age as covariates?

Cheers,

E.J.