Hi,
I just found that the interaction term A*B is not significant (p=1.58). However, in sub-sample analysis,
A is significant when B=0, A is not significant when B =1. How to interpret the result of sub-sample analysis? I think the sub-sample analysis is moderating analysis. However, the interaction term is not significant. Now, I am confused.
I really appreciate any help you can provide.
Dear Kate Xx
These links might be useful, have a look:
1. https://www.statalist.org/forums/forum/general-stata-discussion/general/1432310-how-to-analyze-a-subsample-or-should-i-use-interactions
2. https://parasitesandvectors.biomedcentral.com/articles/10.1186/s13071-019-3606-5
3. https://www.researchgate.net/post/Why_are_the_subsample_results_different_from_the_full_sample_results_in_Linear_dependency_tests
Best Regards
Dr. Fatemeh Khozaei
The finding that one coefficient to be different from zero while the other is not can easily be consistent with them not being different from each other. For example, one subsample might have a small positive effect, while the second has a somewhat larger positive effect, which is enough to make second coefficient different from zero, but not from the first coefficient.
They are different. One take part of the sample while the other use interaction term by multiplying the IV with the DV.
- Badges
- Science topic
- Similar topics
- Counseling Psychology
- Confusion