APS

2022 APS Annual Convention · 2022

Performance of Bayesian Model Fit Indices for Knot Specification in Piecewise Growth Curve Modeling

Chicago, IL · May 2022

Poster · Methodology

  • Lydia Marvin
    University of California, Merced
  • Haiyan Liu
    University of California, Merced
  • Sarah Depaoli
    University of California, Merced

Abstract

Bayesian piecewise linear growth modeling is a flexible tool for capturing nonlinear change. It breaks the overall growth trajectory into connected linear segments. In this study, we evaluated Bayesian model indices for specifying changepoints. Our results suggest the BIC and DIC have decent selection rates for the true model.

Other

← Poster Session VII