Dear all,
I am a student currently engaged in a meta-analysis aimed at exploring the relationship between two variables, A and B. In my review of the literature, I have come across studies that have conducted separate correlation analyses between subscales of A (e.g., A1 and A2) and B, while other studies have compared A and B as a whole.
I am interested in including both types of studies in my analysis. However, I am uncertain about the feasibility and appropriateness of converting the correlations from the studies that analyzed the subscales separately into an overall correlation for comparison with the studies that analyzed A and B as a whole.
Is it reasonable to transform these separate subscale correlations into an overall correlation for A and B for meta-analysis purposes? If yes, then how to achieve that goal?
I would be deeply grateful for any advice or recommendations on a method or approach that could achieve my idea. Any guidance or references to relevant literature would be immensely appreciated.
Warmest regards
Hi,
yes that is possible as long as to inform the model about the nestedness of the data. The key analysis would be a three-level random effects meta-analysis (see my profile for a bunch of applications.
All the best,
Holger
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training