The IES project, Propensity Score Analysis of Multilevel Data: A Systematic Review, will synthesize how propensity score methods have been used with multilevel educational data and translate the methodological literature into an accessible, step-by-step analysis and reporting framework.
Wei Li serves as a Co-Principal Investigator on this project, working with Principal Investigator Walter Leite and Co-Principal Investigator Huibin Zhang.
Within the lab, this project connects causal inference for observational studies, multilevel data analysis, systematic-review methods, and practical guidance for education researchers. It complements the lab’s work on power and sample-size planning for propensity score-based quasi-experimental studies.
Project Information
- Funding agency: U.S. Department of Education, Institute of Education Sciences (IES), National Center for Education Research (NCER)
- Program: Statistical and Research Methodology in Education
- Project type: Methodological Innovation
- Award number: R305D260024
- Principal Investigator: Walter Leite
- Co-Principal Investigators: Huibin Zhang and Wei Li
- Wei Li’s role: Co-Principal Investigator
- Institution: University of Florida
- Project period: August 1, 2026-July 31, 2028
- Award amount: $348,917
Project Goals
- Conduct a systematic review of education studies published from 2005 through 2025 that used propensity score analysis with multilevel data.
- Document methodological choices across data preparation, propensity score estimation, implementation, balance evaluation, treatment-effect estimation, and sensitivity analysis.
- Compare common practices with recommendations from the methodological literature.
- Develop practical guidance that helps researchers select, implement, and report propensity score analyses for nested educational data.
- Use feedback from doctoral-student user testing to improve the accessibility and usefulness of project products.
Planned Products
- A systematic review paper.
- Conference-based training courses on propensity score analysis with multilevel data.
- Analysis flowcharts and reporting checklists.
- A project website and YouTube video series.