230206 UE EC: Moderation and Mediation analysis (2016S)

This is an interdisciplinary data analysis seminar in workshop form focused on the application of principles of linear modeling in the context of linear regression analysis to exploring questions about mediated (i.e., indirect) and moderated (i.e., interaction) effects. We will spend part of the course talking about partitioning effects into direct and indirect components and how to quantify and test hypotheses about indirect effects, part talking about estimating, testing, and probing interactions in linear models, and part integrating moderation and mediation as “conditional process analysis” by discussing and how to conceptualize and test the contingencies of a mechanism (“moderated mediation”) and whether moderated effects are mediated (mediated moderation). Computer applications will focus on SPSS using off-the-shelf code and various macros that facilitate the analysis. It is assumed that you have taken a course in multiple regression and have done well or are otherwise comfortable with the principles of multiple regression analysis (a review of this principles will be offered in the first class of the course). No knowledge of matrix algebra is required or assumed.