Postsecondary Research Clearinghouse Methodology

We employed a semi-systematic search strategy to identify relevant literature. Two members of the project team reviewed studies to assess relevance, skimming the title, abstract, and full text as needed. Our key sourcing methods included:

• Systematic Reviews: We leveraged existing systematic reviews, such as:
  • Dynarski et al. (2023) on nonfinancial barriers to college access
  • French & Oreopoulos (2017) on behavioral barriers to college transition
  • Herbaut & Geven (2019) on reducing inequalities in higher education
  • Lindsay et al. (2024) on Career and Technical Education (CTE)
• What Works Clearinghouse (WWC): We consulted WWC-reviewed studies.
• Database Searches: We searched Education Resources Information Center (ERIC), Education Source, and Econ Lit using a set of search terms.
• Working Papers: We searched NBER Working Papers and Annenberg Institute’s EdWorkingPapers.
• Research Organization and Government Agency Websites: We reviewed the following set of organizations that conduct rigorous quantitative studies: Abt Associates; American Institutes for Research (AIR); Community College Research Center (CCRC); Mathematica; MDRC; National Center for Education Research (NCER); RAND Corporation;  SRI International; Westat; WestEd
• Backwards / Forward Searches: For key studies, we reviewed the references included in the study as well as subsequent papers that cite the study. We also used tools such as ConnectedPapers and ResearchRabbit to find related literature. 

After sourcing relevant literature, two team members used the criteria below to determine whether the study would be included in the Clearinghouse. Notably, we do not limit the Clearinghouse to only interventions that show effectiveness. Eligibility for the Clearinghouse is based on the rigor of the study design, regardless of the results, meaning studies with null effects are also included. We excluded dissertations from the Clearinghouse. 

• Publication Date: Published in the past 20 years (since 2004).
• Research Setting: Conducted in the U.S.
• Primary Focus: Measured the effectiveness of the intervention / program of interest.
• Research Design: Employed one of the following research designs that had a comparison group and used a research design that addresses selection bias to plausibly estimate the causal effects of an intervention: Randomized Controlled Trial (RCT); Regression Discontinuity (RD); Difference-in-differences (DID); Instrumental Variables (IV) Estimation; Comparative Interrupted Time Series; Fixed or Random Effects (with panel data); Matching Design (e.g., Propensity Score Matching (PSM), Inverse Probability of Treatment Weighting (IPTW)).

For studies that pass the screening stage and are selected for inclusion in the Clearinghouse, five team members extracted key details, such as the intervention description, sample characteristics, and findings. These details were used to create individual study summaries; support synthesis across studies with similar interventions; and help group studies into broader research categories. To streamline the extraction process, we used the AI tool Elicit to complete the first pass at extracting key study details. These extractions were reviewed and revised by our team members to ensure accuracy. Study summaries were then written by research team members, based on the extracted details.