Help with data analysis plan for dissertation

#1
Any help would be much appreciated :)

For my research project I'm hoping to do a moderation analysis and will be looking at if Variable M moderates the relationship between Variable X and Variable Y. All variables are continuous (scale) and were measured at only 1 time point.
My 3 hypotheses are:
1. Variable X predicts Variable Y
2. Both Variable X and Variable M predict Variable Y
3. Variable X and M interact to predict Variable Y (moderation)

I'm planning to use Hayes' PROCESS Macro to run the moderation analysis on SPSS but just wanted to know if there's anything I needed to do beforehand i.e run a multiple regression to check assumptions? Spent a good deal of time looking through Andy Field's book but am still a bit confused! :confused:
 

staassis

Active Member
#2
To the best of my knowledge, the PROCESS add-on assumes normally distributed residuals. If they are not normally distributed and the sample size is modest, you are better off estimating various models with interactions using standard SPSS functionality (standard SPSS edition). In many "linear" and "non-linear" sub-menus there, you will see the "Bootstrap" button. This will lead to more accurate statistical inference.

In general, be cautious about statistical software which was not developed by statistics experts but rather by social and medical scientists.
 
Last edited:

hlsmith

Less is more. Stay pure. Stay poor.
#3
To me, moderation represents interactions not mediation. Are you referencing moderation or mediation or moderated-mediation? You really don't need a macro or anything special to examine moderation on the multiplicative scale. When examining for it on the additive scale a little extra coding my be necessary.