1
$\begingroup$

I have conducted a simple moderation analysis in SPSS using the PROCESS macro and I've found that the interaction XM is not significative (p=0.87), but almost all three conditional effects of X on Y at values of the moderator are significative:

  • Moderator=1.64 Effect=1,17 p=.00 95% CI [0.38 - 1,97]
  • Moderator=2.58 Effect=1,22 p=.00 95% CI [0.66 - 1,79]
  • Moderator=3.52 Effect=1,27 p=.00 95% CI [0.45 - 2,09]

The same happens with the Johnson-Neyman technique, as I see that Moderator value(s) defining Johnson-Neyman significance region(s):

Value= 4.50 % below=98.44 % above=1.56

I'm truly lost. Can anybody help me with the interpretation of these results? How can it be that all the conditional effects of X on Y at values of the moderator are significant, but the interaction XM is not?

$\endgroup$
1
$\begingroup$

The interaction is the degree to which the effect of the focal predictor changes based on different levels of the moderator. The focal predictor might have a significant effect at all levels of the moderator; in fact, this is what you see. But the effect of the focal predictor may not vary much at different levels of the moderator. In your example, for large changes in the moderator, the effect of the focal predictor changes very little.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.