When: Monday, Nov 27, 2017 @ 10:00 AM – 11:30 AM

Where:  Norm Endler Seminar Room (BSB 164)

Speakers: Joo Ann Lee, York University

Title: A comparison of methods in modelling nonlinear data

Abstract: One of the fundamental assumptions of the General Linear Model (GLM) is linearity in the effects or relationships between the variables. We often hear passing remarks such as “in real-world data, the effects may not be linear” from critics of practitioners relying mainly on the GLM in modelling their data. Assuming these critics are correct, what are the potential downfall of using the GLM to model nonlinear data? Moreover, how well do other (and newer) proposed methods such as Regression Trees do in modelling nonlinear data? This talk investigates these issues via a simulation study of nonlinear data situations often encountered in psychological science.