I conducted moderation with age as a moderator, using PROCESS. the age range is 65-91. the program cut the age groups as follows: 67, 74, 82. what does it actually mean? If this means: group 1: participants between 67-74, group 2 =74 -82, group 3 above 82, then the participants between 65-67 were not included? (only 3 groups were created).Thank you
@Cohen
There are two methods for segregation of the moderator in PROCESS Macro. One is -1 SD +1 other is First and Third Quartile method. In your case, if the interaction effect is negative, it refers to negative moderation and means that the relationship between X and Y variables reduces if age increases and vice-versa in case of positive moderation effect.
It is also possible to perform using the free JAMOVI program, which is adapted to the moderated mediation model based on the factorization of the moderated variable according to its category.
Miri Cohen there is one very important thing, which has not been stated clearly so far: whatever method is used (+/- SD, Percentiles or own values) it DOES NOT create groups!!! This are just "pick a point" methods where you calculate the effect of X on Y at this particular values of your moderator W. That's why it is called "simple slope analysis", because you chose points (how many you like) and calculate the slope of X for this specific point. There is no grouping!
The +/- 1SD "rule" (including typically also the mean/median of W as middle reference) was employed to show slopes for participants which have low and high values. This is nice, but may lead to impossible values, if W is skewed. Therfore, nowadays 16th, 50th and 84th percentile are preferred since it guarantees that you have actual values which are part of the distrinution of W and in case of normality it will show the same values as +/- 1 SD. If you have any interesting or relevant cut off points (maybe due to a theory behind it) you should use them at least additionally.
Another approach, which is not pick a point, is to use the Johnson-Neyman approach, which calculates the slopes for the whole range of values of W, so you can see the transition from significant to non significant slopes of X.
- Badges
- Science topic
- Similar topics
- Computer Science
- Program