Full course description
Term: Fall 2020
Date: November 6th, 2020
Time: 12:30pm - 1:30pm
Location: ONLINE ONLY
Instructors: Kevin McKee, Jennifer Van Mullekom, & Kate Miller
Presented By: Statisical Applications and Innovations Group (SAIG)
Mediation and moderation models are often used to draw inferences about indirect causation. This presentation shows both can be understood through the broader context of structural equation models and analyzed intuitively through the graphical language of path tracing. Convenient strategies are demonstrated for including categorical variables, counts, and Likert scales anywhere in a linear model. Conceptual interpretation of fit statistics and requirements for causal inference are further considered.
About the Speaker: Kevin L. McKee, Ph.D. earned his multi-disciplinary doctorate in Psychiatric, Behavioral, and Statistical Genetics at Virginia Commonwealth University in Richmond, VA. His dissertation research focused on methodology for psychological time series and the use of genetic signals to validate theories of cognitive and behavioral processes. He specializes in using R statistical language for time series analysis, simulation, power calculation, and structural equation modeling with software such as OpenMx. Kevin has contributed to a variety of domains including algorithms and methodology, analysis of high-frequency sensor data to study anxiety and panic disorders, self-report from experience sampling studies of substance use, neuroimaging in psychiatry, and epidemiological survey data.