|Title:||Integrated Replication Designs for Identifying Generalizability Boundaries of Causal Effects|
|Principal Investigator:||Wong, Vivian||Awardee:||University of Virginia|
|Program:||Statistical and Research Methodology in Education [Program Details]|
|Award Period:||3 years (09/1/2022 – 08/31/2025)||Award Amount:||$899,115|
|Type:||Methodological Innovation||Award Number:||R305D220034|
Co-Principal Investigator: Steiner, Peter M.
The purpose of this grant is to develop an approach for identifying generalizability boundaries, which describe conditions under which effects are expected to replicate across variations in units, treatments, outcomes, settings, and times. The researchers will use principles of fractional and confounded factorial designs to plan integrated fractional replication designs and use subject-matter theory to specify causal estimands of interest, along with hypothesized moderators. The research will proceed in three phases, ultimately yielding user-friendly software for running the models, workshops and presentations at conferences, and papers in peer-reviewed journals.
First, the team will conduct Monte Carlo simulation studies to investigate the effects of different design facets on the results from integrated fractional factorial replication designs. The simulations will also be used to test the robustness of the results to deviations from design assumptions. Second, the researchers will demonstrate the use of the models using real data from special education and from teacher preparation settings. In the third phase, the research team will create a user-friendly version of the software in R, along with instructional materials for implementing and analyzing integrated fractional replication designs. The instructional materials will be used at conference workshops and will be available online.
Related IES projects: Developing Methodological Foundations for Replication Sciences (R305D190043); Developing Infrastructure and Procedures for the Special Education Research Accelerator (R324U190001)