I would like to perform different moderator analyses to test if a pornography consumption index can predict sexist attitudes and if that relationship is eventually moderated by gender.
The pornography consumption index is a continuous variable computed by summing up four z-standardised interval scaled measures. Gender is a binary variable (0 = female, 1 = male).
As far as I know centering is done to make the lower-order effects (pornography consumption, gender) interpretable, since in a lot of cases a 0 value on one of the predictors isn't interpretable. In my setting a 0 is interpretable for gender (gender: 0 = female) and the pornography index is already centered, since the four variables it results from were z-standardized.
Can I conclude that I don't need to center the predictors of my model in that case?
Thank you in advance!
If I remember correctly, the short answer is yes.
On place to check would be here:
Article Moderation in Management Research: What, Why, When, and How
Heiko Breitsohl Thank you, this article was really helpful! Categorical moderators, that are binary, do not need to be centered. But still I am not sure, what to do with the pornography index. According to the article it would be senseful to center it, but as I wrote earlier, it is the sum of four z-standardized variables and I am unsure if it is already centered in that case?
Well, what is keeping you from simply computing a mean score of the unstandardized variables, and then centering that?
By the way, if those for variables are based on a factor model, it might be a good idea to use structural equation modeling.
The four items had different likert scales and should have an equal influence on the index. Therefore I used the z-standardization and summed them subsequently. This index is based on Hald (2006, 2013), who could show that these four items can be combined as one measure quantifying the level of pornography consumption in the past.
Unfortunedly I am totally unfamiliar with structural equation modeling.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS