Hello Annabel,

The negative sign for the interaction of moderator and IV derives from the fact that, for level 0 of the "cultcat" variable, the effect (1.3465) is higher than the effect for level 1 of the cultcat variable (0.7054). So, there's an inverse relationship between cultcat level and resultant effect. (Can't be really specific without knowing what your variables were!)

You might wish to plot simple slopes, or plot means for the various combinations as a visual aid in portraying (and explaining) your results.

Good luck with your work.

David Morse

Hi David,

Thank you for your quick and helpful response.

My moderation variable is culture (culcat) which i categorised into, Individualism = 0 and Collectivism = 1. My dependant variable is consumption and the independent is entertainment. From the numbers I gathered, I assumed that the results on a graph would show a negative relationship, but from what I can see it appears positive. I have attached a picture to help my explanation. Maybe I am missing something...!

Thanks,

Annabel

Hello again, Annabel,

Note that, when the culture variable equals 0 (individualism), the estimated score on ncontot is generally higher than when the culture variable equals 1 (collectivism). That's an inverse relationship. Don't conflate that with the positive relationship between enttot and ncontot.

Good luck finishing your project.