Read on Twitter
1/7. I reviewed the literature on what the evidence (to date) shows on tests and test optional policies in college admissions as part of a Q&A panel for @ConversationEDU (w/Angela Farmer): https://theconversation.com/many-colleges-have-gone-test-optional-heres-how-that-could-change-the-way-students-are-admitted-144998 (thanks to @dcwriter360 for editing)
2/7. Here I highlight some core articles. The evidence supporting the use of tests in admissions is quite substantial though often completely ignored in public debates. For example, Kuncel and colleagues: https://www.researchgate.net/publication/8922929_Academic_Performance_Career_Potential_Creativity_and_Job_Performance_Can_One_Construct_Predict_Them_All
3/7. An analysis of Bowdoin, which has been test optional since 1969, found that those who didn’t submit their SAT scores had lower SAT scores, and that they did worse during their first year of college compared to those who submitted their scores. Wainer: https://blog.criteriacorp.com/dont-ask-dont-tell-the-new-rules-of-the-sat-and-college-admissions/
4/7. Test optional shifts weight to other criteria. Grades are susceptible to inflation. Letters of recommendation are weakly associated with student success: https://onlinelibrary.wiley.com/doi/abs/10.1111/ijsa.12060. For more research on test optional consequences see Saboe and Terrizzi: https://www.sciencedirect.com/science/article/abs/pii/S0165176518304300?via%3Dihub
5/7. Research is mixed on whether there’s a long-term payoff associated with school selectivity. Dale and Krueger showed ( https://www.nber.org/papers/w7322 ) going to a highly selective college does not appear to matter for long-run income, but that SAT scores do ( https://psyarxiv.com/rpgea/ )
6/7. More research and discussion on returns to a more selective college from Hoxby ( http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.197.6607), Chetty et al ( https://www.nber.org/papers/w23618 ), and further discussion ( https://www.econlib.org/archives/2013/11/hoxby_vs_dale-k.html)
7/7. In sum, there is a very large literature in support of the predictive validity of tests and use in admissions. There is a small amount of literature on test optional policies, but that research is mixed. Future research on test optional or removal would be useful.
