Answers are grounded in this paper and cite their sources.
The appropriate interpretation of interaction terms in studies employing moderated multiple regression ( MMR) analysis is considered. It is argued that given a theory-based, a priori hypothesis, an MMR analysis is analogous to a planned statistical comparison and, thus, a significant overall F value is not a prerequisite for interpreting a significant interaction term.
No takes yet. Share an insight, caveat, or question.
Bedeian et al. (1994) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: