The Psychology department is hosting a colloquium by W&M alum Dr. Kris Preacher, Professor and Associate Chair of Psychology at Vanderbilt, on March 20, from noon-1pm in Tidewater B, Sadler Center. Kris is going to be presenting his recent work on mediation analyses, including the various methods and underlying assumptions for testing indirect effects in applied settings (i.e., real data, not simulations). He will also illustrate how to use online tools (and other software) he's developed to perform these tests easily and accurately. For more information, please see his website (quantpsy.org) and abstract below.
His talk will be geared toward students and faculty. If you have any students working as RAs, doing independent studies, or in research methods (or statistics) courses, please encourage them to come! Graduate students and post-doctoral students are also encouraged to attend.
We expect the presentation will be relevant across departments/schools: psychology, economics, biology, sociology, education, chemistry, anthropology, among many others. Please circulate to potentially interested colleagues. For more info, or to arrange a one-on-one meeting with Kris, email firstname.lastname@example.org.
Summary of Dr. Kris Preacher's Talk:
Indirect effects: Inference and modeling extensions
I will begin with an overview of basic mediation analysis, followed by a consideration of modern methods of significance testing and confidence interval construction for mediated, or indirect, effects. The second half of the talk will cover extensions of mediation models to address important research questions often raised by social scientists. These extensions include: multiple mediator models, moderated mediation, nonlinear indirect effects, longitudinal mediation, and mediation in multilevel (or clustered) data. I will conclude by discussing future directions in mediation research.