On the importance of temporal precedence in mediational analyses.

Abstract

This response to the commentary by Gaynor (in press) (a) underscores the importance of defining terms (predictor, moderator, and mediator variables) used in psychological treatment outcome research, (b) notes the improvements in mediation analyses using treatment of youth anxiety as the illustration, and (c) encourages continued advancement in the methodology for mediation analyses. We agree with Gaynor (in press) both on the importance of temporal precedence and on the use of cautious language with regard to interpreting mediational analyses. (PsycINFO Database Record