EdWorkingPapers posted an RCT of three STEM summer programs for underrepresented high schoolers. Quick take: The findings touted in the study abstract — including large impacts on college STEM degree receipt — are quasi-experimental, not RCT. The RCT impacts are mostly small & not statistically significant.

• The headline finding that 1 of the programs (the 6-week version) increased college STEM degree receipt by 33% is a quasi-experimental (non-RCT) comparison of higher-ranked vs lower-ranked students, with statistical adjustments to try to equate the two groups.

• The study abstract - shown below - should make this clear (not just say it's "RCT"), as such quasi-experimental methods are not as credible as high-quality RCT designs.

• The paper does report RCT results for 1 of the programs (the online version). The effects found on STEM degree receipt, college graduation, & other outcomes at the 4 year mark are small and not statistically significant.