Read on Twitter
Some musings about graduate admissions, standardized testing, etc while I wait for my coffee.... these are things that seem obvious? but are maybe not obvious?
1. any proposal to get rid of X is simultaneously & necessarily a proposal to use Y instead. If you drop the GRE, for instance, what are you using instead?
2. If you believe, as I do, that educational systems are profoundly unequal from early childhood, then *all* indicators of academic skills will show group differences by the time people are in their 20s
3. Doing something might *feel* better than doing nothing, but feeling better is not an excuse for policies that make equity problems worse instead of better
4. It is possible to make a change that fails to improve race equity, while also reducing efficiency in the allocation of scarce educational resources and reducing other equities we care about (eg, social class)
5. Trying to come up with a definition of “merit” that gets at what people “deserve” and is free of morally arbitrary luck is a fools errand. We can’t do it and we wouldn’t want to if we could.
6. The lesson of social science is that our intuitions about what makes a system work better are frequently BAD. This is why we do science.
7. Given lifelong inequalities in education, damaging effects of poverty, etc, the search for a graduate admissions indicator that is unrelated to family background implicitly relies on notions of “inherent” talent which I personally find a bit creepy
8. Examining associations between any criterion and a focal outcome in samples of people who have already been strongly selected on the criterion is statistical garbage and y’all should know better
9. Predicting complex social outcomes like “scientific productivity” is really fucking hard.
10. If we throw out correlations of .1 - .3 as meaningless we are throwing out most of social science, including most of the studies y’all cite to show that poverty matters for understanding child development
11. Changing the composition of the 2% of Americans who have PhDs might be an important goal but it is a really really really small rudder for the larger ship of economic inequality
12. Many people have not asked themselves whether they care about equity in higher education as a END versus as a MEANS to equity in some other more basic good
13. The emphasis on equity in higher education as a means to equity in income and other goods reinforces the meritocratic narrative of people “earning” and “rising”. Many people have not asked themselves if this is a narrative they want to buy into.
14. All of the conversations about educational selection would be less fraught if we actually cared about the dignity of all labor in this country.
<pause> my coffee and tacos are ready.... might resume later
15. Work samples are generally the best predictor of job performance. If departments were serious about revising admissions they could start there.
16. Undergrad GPA is a very approximate rating of four years’ worth of work samples, but only captures part of the type of work done in grad school in the sciences
17. Generally speaking, *expanding* standardized testing to be more universal increases the representation of Black and low-income students
18. My bet is that making the GRE free, making every junior in college take the GRE would expand the # of students from historically excluded groups in PhD programs
19. The incentives of faculty who are making admission decisions are frequently not aligned with the department’s equity goals
20. It is generally a bad idea to stake moral claims for equity & inclusion to easily disprovable empirical claims about test bias
Ok gonna stop there, because 20 is a nice round number and I have to write more recommendation letters.
