I am working on PhD proposal. I had to work with four independent variables with some moderating and mediating as well outcome variables. If I only manipulate two out of four independent variables while conducting lab exp in Psychology and do the rest with empirical study. Should it would be an acceptable practice in psychology research?
kind regards
Dear Muhammad,
Acceptable in the sense of typical scientific practice, yes. Advisable, no.
Four independent variables are nearly always too many. It hurts you twice. First, you spread the data thinly (2^4=16 or worse) and you have to do intensive multiple corrections. To perform a well-powered study under these conditions is most often impossible.
Can you fix the other independent variables? Or match them between the other groups (e.g. age matched controls)? Or have a well motivated theoretical model for those variables? Try to get rid of the additional independent variables. If all else fails, then do a hierarchical model.
Best, Peter
- Badges
- Science topic
- Similar topics
- Psychology