What is the best way to compare two non-homogeinous populations (Levene's) with covariates (e.g. age). I know ancova assumes homogeneity. Is there a way around this? Many thanks
Asked
Active
Viewed 289 times
0
-
Somewhere around the nonsense word "compacovary" it appears some of your original text was lost during the posting of this message. Please consider editing it to reflect what you originally intended. – whuber Mar 07 '19 at 17:10
-
Thanks and apologies. Edited to reflect the question accurately. – hH1sG0n3 Mar 07 '19 at 18:37
-
This is rather broad. You can edit again if you want to try to refine it. That said, it appears to be a duplicate of [Alternatives to one-way ANOVA for heteroskedastic data](https://stats.stackexchange.com/q/91872/7290). That question is asked in terms of a one-way ANOVA, but one-way ANOVA vs factorial ANOVA vs ANCOVA largely doesn't matter: most of what's written there will apply here. If you are asking about something else, you need to narrow down & clarify your question. – gung - Reinstate Monica Mar 07 '19 at 19:46