|Title:||Assessing Generalizability and Variability of Single-Case Design Effect Sizes Using Multilevel Modeling Including Moderators|
|Principal Investigator:||Moeyaert, Mariola||Awardee:||State University of New York (SUNY), Albany|
|Program:||Statistical and Research Methodology in Education–Early Career [Program Details]|
|Award Period:||2 Years (07/01/19–06/30/21)||Award Amount:||$224,997|
|Type:||Methodological Innovation||Award Number:||R305D190022|
The increasing number of published single-case experimental design (SCED) studies in education sciences can be used to inform policy, research and practice decisions. Such decisions, with a large impact, should be based on scientific knowledge. For this purpose, large bodies of literature in the SCED field can be summarized in a standardized, objective, reliable and valid manner. One technique that is developed and can be applied to serve this purpose is multilevel meta-analysis. Because in a single SCED study multiple cases can be involved and for each case an effect size is calculated, A three-level meta-analysis is recommended as this takes the hierarchical structure of the SCED data into account: namely, effect sizes are clustered within cases and cases in turn are clustered within studies.
The goal of this research project is to contribute to evidence-based decisions, research and practice in education through designing multilevel meta-analysis. The research team will empirically validate the multilevel meta-analytic model including moderators in order to explain variability among effect sizes at the case and at the study level. They will also develop and empirically validate power calculations to detect meaningful moderator effects. To accomplish these validation goals, the researchers will conduct large scale Monte Carlo simulation studies so as to provide guidance for study design. Finally, researchers will evaluate the current What Works Clearinghouse (WWC) standards for combining SCED studies (including moderators) and make recommendations based on the results of the empirical validations described above.
Related IES Projects: Multilevel Modeling of Single-subject Experimental Data: Handling Data and Design Complexities (R305D150007)