In order to determine which interventions work best for which students, precision education researchers have examined aptitude-treatment interactions (ATI) or skill-by-treatment interactions (STI) by using multiple regression analysis. In recent years, technological innovation and greater accessibility to more advanced statistical techniques have allowed researchers to examine ATI or STI using longitudinal mixed modeling. Probing techniques for interaction effects like slopes difference tests fit well with an ATI or STI framework, but the power of using a slopes difference test in longitudinal mixed models is unknown. The current study used the simulation method to determine which factors influence the power of a slopes difference test in a longitudinal mixed model. Specifically, the study examined design conditions such as number of waves, number of clusters and participants per cluster, effect size, intraclass correlation, variances and covariance of the random effects, and proportion of treatment to non-treatment participants. These were examined in order to help ATI and STI researchers understand whether the slopes difference test in a longitudinal mixed model is sufficiently powered for their sample and study. Effect size, number of waves, number of clusters, and participants per cluster were found to be strong determinants of power and significantly impacted model convergence and rates of singularity. The slopes difference test was also found to have greater power in a longitudinal mixed model than was discovered for a cross-sectional multiple regression context.
| Identifer | oai:union.ndltd.org:unt.edu/info:ark/67531/metadc1833466 |
| Date | 08 1900 |
| Creators | DeJong, Trey L. |
| Contributors | Chen, Qi (Educational psychologist), Boesch, Miriam C., Henson, Robin K. (Robin Kyle), Zhang, Tao, 1978- |
| Publisher | University of North Texas |
| Source Sets | University of North Texas |
| Language | English |
| Detected Language | English |
| Type | Thesis or Dissertation |
| Format | vii, 62 pages : illustrations, Text |
| Rights | Public, DeJong, Trey L., Copyright, Copyright is held by the author, unless otherwise noted. All rights Reserved. |
Page generated in 0.0017 seconds