Time-Specific Errors in Multilevel Growth Curve Modeling in Educational Research (Poster 46)

• In Event: Sunday Poster Session 1:15 pm
  In Poster Session: Educational Statisticians SIG Poster Session

Abstract

In longitudinal data, there are often circumstances that are specific to time, called time-specific errors, that may bias standard error estimates and thus result in incorrect inferences (Usami & Murayama, 2018). This study examines the influence of time-specific errors in multilevel growth curve models (GCMs) for academic achievement outcomes. Parameter and standard error estimates are compared between the linear and quadratic GCMs with and without time-specific errors specified. The inclusion of time-specific errors impacted standard error estimates for the intercept and time variables, which highlights the importance of appropriate model specification when analyzing longitudinal data as model misspecification may result in different conclusions of the growth functional form.

Authors

• Sungeun Lee, Georgia State University

• Audrey J. Leroux, Georgia Institute of Technology

• Kevin Fortner, Georgia State University