Project Activities
The project team will perform simulation studies of two methods for estimating site-specific effects and effect-size distributions in multisite trials, and then build an R package and an R Shiny app to enable modeling using these methods while applying the resulting best practices. The team will also release a practice guide documenting the approach.
Structured Abstract
Research design and methods
This project will translate the insights gained from the simulation studies into a set of comprehensive guidelines and practical tools. Simulations will be calibrated to reflect typical data scenarios encountered in multisite trials of education interventions.
User Testing: The project team will conduct an extensive testing and validation phase, using the expertise of an advisory panel comprising researchers in the field of education statistics. End-user testing will be conducted across multiple education research settings, including universities and policy analysis offices, to ensure the usability and applicability of the developed products.
Use in Applied Education Research: By providing education researchers with improved tools and guidelines, this project team aims to enhance the quality of multisite trial analyses, leading to more accurate and reliable interpretations of site-specific effects. The products developed will contribute significantly to the field of educational research, offering researchers practical solutions to overcome existing challenges in multisite trial analyses.
People and institutions involved
IES program contact(s)
Products and publications
Products: The outcomes of the simulation study will inform the development of three key products:
- Comprehensive Guide for Applied Researchers (Product 1): A detailed guide providing step-by-step recommendations for analyzing multisite trial data. It will be grounded in empirical findings from the simulation study, supplemented with real-world case studies.
- Interactive Shiny Web Application for Analytic Decision-Making (Product 2): An online tool enabling researchers to simulate and visualize the performance of different model-method combinations in various scenarios, aiding in the selection of the most effective strategies for their specific research settings.
- Streamlined R Package for Optimal Bayesian Methods (Product 3): A user-friendly R package to facilitate the implementation of recommended methodologies in analyzing multisite trial data, employing various Bayesian models and methods.
The project team will disseminate these products via strategy workshops, virtual learning modules, participation in academic conferences, and publications in educational research journals. This multifaceted approach aims to reach a wide range of audiences, including applied education researchers, methods researchers in schools of education, and education policymakers and administrators.
ERIC Citations: Find available citations in ERIC for this award here.
Related projects
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
