PI: Dr. John Ferron, Education

Sponsor: IES

Series of statistical simulations studied design to examine the usefulness of multilevel models to analyze and meta-analyze data obtained from a variety of single-subject experimental designs (SSED). In the proposed project, we want to study data complexities that are likely to occur when analyzing real SSED data. In the project, the complexities that we will address will be looked at from four interrelated perspectives: (a) the way(s) the complexities can be translated or accounted for in the statistical models, (b) the estimation of the model parameters, which might require more complex estimation procedures including Bayesian and bootstrapping estimation, (c) the way data should be standardized, (d) effect size metrics that can be used to summarize effects for individual participants and studies. We will also look at statistical consequences of the practices of applied researchers, this is, the way they collect the data or set up the study.