Hi! I am doing a moderation analysis using model 1 in PROCESS by Hayes. I have a binary IV, and the output tells me my moderator is not significant. however, I would like to check if it is significant or not for each conditions of my IV.. But the output is not generating the conditional effects, I have tried everything and nothing works. What should I do?
thank you!
Hello Francisca,
If the moderation (e.g., IV, "X," by W interaction, where W is the purported moderator) is nonsigificant, then that implies there are no simple/conditional effects that differ from one another.
If you wish to verify this, you could run the simple effects tests (IV differences at a specific level of W; and, for that matter, W differences at specific levels of the IV). These are the conditional tests.
In SPSS, the syntax to elicit simple effects tests for a two-way design would look like this:
unianova Y by X W
/ method = sstype(3)
/ intercept = include
/ emmeans = tables(X*W) compare (X) adj(bonferroni)
/ emmeans = tables(X*W) compare (W) adj(bonferroni)
/ print = descriptive
/ criteria = alpha(.05)
/ design = X W X*W .
Where: Y = DV; X = IV; W = moderator.
The two lines starting with "emmeans" are the commands for the simple effects/conditional tests.
Good luck with your work.
