Hi!
I have been looking at my data and got interested in covariates. Now, when exploring covariates η2 changes depending on which covariates I input. For example cavariate 1 has a noticeable η2 but when I add covariate 2 and covariate 3 into the model all η2s are not significant or really small. How can I then interpret the high η2 of covariate 1 when it is in the model alone (especially if the effect disappears when I input other covariates)? Also, is it even valid to report this effect when it disappears if I add more covariates in the model? Would it be valid to do independent ANCOVAs for each covariate?
Looking forward to your insights!
Best,
Anna