Page 1 of 1

Comparing multiple mean differences

Posted: Mon Aug 16, 2021 9:09 pm
by MM15
How would you compare multiple mean differences?
Anova post hoc tests Show significant differences for targets and non targets in different emotion categories . Now i want to know if these differences are significantly bigger for one Emotion category compared to another

Re: Comparing multiple mean differences

Posted: Mon Aug 16, 2021 11:18 pm
by jonathon
so the post-hoc tests (ostensibly) tell you if there's a significant difference between groups ... now you're wanting to test which difference between groups is greater?

i think you'll need a linear mixed model for this ... see gamlj.

jonathon

Re: Comparing multiple mean differences

Posted: Tue Aug 24, 2021 4:39 am
by reason180
Given Emotion A and Emotion B, a 2-by-2 ANOVA will tell you whether the effect of Target vs. Non-Target is significantly different for Emotion A than for Emotion B. Specifically, if there is a significant interaction, then the difference between 54 and 28 is significantly different than the difference between 93 and 71.

Example pattern:
M = 54 [Non-Target; Emotion A]
M = 28 [Target; Emotion A]
M = 93 [Non-Target; Emotion B]
M = 71 [Target; Emotion B]