Welcome to Generalizability.org!

Impact studies are conducted to inform decisions by policymakers and practitioners. But research findings do not always apply beyond the samples that produced them. “Generalizability” or “external validity” captures the relevance of research findings to a broader group of people or places. “Transportability” captures their relevance to a different group of people or places.

This website provides resources for researchers aiming to conduct impact studies that are more generalizable, externally valid, or transportable. These resources include:

• Publications. We provide a library of information about journal articles and other publications that develop, demonstrate, summarize, and test the performance of statistical methods for assessing or improving the generalizability of impact evaluations. This library also includes information about other papers on generalizability, external validity, or transportability that may interest researchers conducting or analyzing data from impact studies.

• Software and code. We provide links to existing software or tools designed to support researchers in conducting more generalizable impact studies and in using existing data to enhance the generalizability of studies. We also offer new code for selecting representative samples from populations of potential study sites.

These resources were assembled by Rob Olsen (George Washington University) and Elizabeth Stuart (Johns Hopkins University), with support from other researchers at Johns Hopkins University, Abt Associates, and Westat.

We acknowledge funding from the Institute of Education Sciences (Award # R305D190020). Any errors on this website are ours alone.

Have questions about available resources? Or suggestions of resources to add to this repository? Please contact Rob Olsen (robolsen@gwu.edu) or Elizabeth Stuart (estuart@jhu.edu).