changing η2s in exploratory ANCOVAs

Discuss statistics related things
Post Reply
annathej88
Posts: 1
Joined: Mon Jul 18, 2022 4:45 am

changing η2s in exploratory ANCOVAs

Post by annathej88 »

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
Post Reply