Multivariate distance matrix regression (MDMR) and multivariate tree boosting (MVTB) are tools designed to detect and describe relationships between predictors and multivariate outcomes, which are common in behavioral research. The traditional way to explain structure in a multivariate outcome is to invoke the linear model (i.e., multivariate multiple regression, MANOVA), but the simplifying assumptions underlying this approach are often inappropriate in the social sciences due to the complexity of human behavior. MDMR and MVTB are alternative approaches that relax these assumptions to varying degrees in order to more flexibly model complex multivariate outcomes.
In this practice job talk, I will motivate and discuss my research on MDMR and MVTB with a focus on their potential value in the context of behavioral research. I am focusing my job search on departments comprised of both quantitative and applied behavioral researchers, so this talk is designed to cater to a broad audience. I would appreciate as much substantive and stylistic feedback as the audience is comfortable providing, and I would also be happy to discuss any suggestions in person or via email at a later date as well. Thank you in advance!