Evaluating Subject Mobility in Partial Nesting Contexts: Multiple Membership Partially Nested Models

• In Event: Innovative Statistical Methods and Computational Tools for Complex Research Designs

Abstract

This study proposes a model to take into account multiple membership partially nested (MM-PN) data, where some level-1 individuals are nested within level-2 groups and other level-1 individuals remain un-nested. Furthermore, a portion of the nested individuals also exhibit mobility among level-2 groups. The estimation of the proposed MM-PN model is compared to a partially nested (PN) model that ignores mobility. This modeling approach was demonstrated using a subset of the Early Childhood Longitudinal Study, Class of 2010-11 dataset to compare homeschooled and traditionally schooled students’ science achievement. Results revealed that parameter and SE estimates were consistent, with the exception of the level-2 variance component and level-2 covariate’s coefficient resulting in differences between the MM-PN model and the PN model.
 

Authors

• Yara Al-nouri, Georgia Institute of Technology

• Audrey J. Leroux, Georgia Institute of Technology