My oldest child is a senior in high school. In applying to schools, he went by some internet wisdom that suggested he only submit his test score if it was above the mean for that school. I wonder if that was just an invitation for them to assume the worst. His results are not all back yet. Is there any data that controls for GPA? The claim I've heard is that standardized tests are bad (for various reasons) and that GPA is the better predictor of success anyway. So are the schools actually weighting GPA more for no-test applicants? Or are they just punishing the no-test applicants? I fear that the actual answer is that they punish the no-test applicants that whose identities they are not interested in, and reward those whose identities are more desirable. I suspect the data to sort that out would be very hard to come by.
Hi Dennis! I offer these thoughts for your younger kids -- not at all as a criticism of you or your son.
1. I think the 25th and 75th percentiles for test scores give a better idea of the distribution than the mean.
2. An applicant might decide not to "spend" one of his 12 or so applications on a college where his test scores are closer to or lower than the 25th percentile score -- especially if he has no other particularly appealing trait for colleges (legacy, relative of donor, athlete, not white or Asian).
3. You're right that there is no publicly available data that tell us the marginal effect of not submitting test scores. The best data we have come from the Students for Fair Admission lawsuits against Harvard and UNC, which cover a period before test-optional policies were common. Litigation pried it out of them.
4. Even ostensibly test-optional colleges let in score submitters at higher rates. Why?
a. One possible reason is that intelligence is the lurking variable here. Smarter kids get higher test scores, which they are then proud to send in, AND exhibit other traits that admission officers like (e.g. lots of AP classes).
b. Related: your intuition that colleges assume low test scores for non-score submitters is probably right.
c. An SAT score is like a credit rating from S&P. Non-score submitters are riskier bets (e.g. "will this kid wash out and pull down our 4-yr graduation rate?"), because colleges know less about them. GPAs reflect both grade inflation and ability, whereas standardized tests are standardized.
5. Test-optional policies probably do enable colleges to let in some low-scoring applicants who belong to the favored groups listed above without those scores pulling down the average test score for the US News ranking.
Please let me know if there's anything else I can do for your family!
My oldest child is a senior in high school. In applying to schools, he went by some internet wisdom that suggested he only submit his test score if it was above the mean for that school. I wonder if that was just an invitation for them to assume the worst. His results are not all back yet. Is there any data that controls for GPA? The claim I've heard is that standardized tests are bad (for various reasons) and that GPA is the better predictor of success anyway. So are the schools actually weighting GPA more for no-test applicants? Or are they just punishing the no-test applicants? I fear that the actual answer is that they punish the no-test applicants that whose identities they are not interested in, and reward those whose identities are more desirable. I suspect the data to sort that out would be very hard to come by.
Hi Dennis! I offer these thoughts for your younger kids -- not at all as a criticism of you or your son.
1. I think the 25th and 75th percentiles for test scores give a better idea of the distribution than the mean.
2. An applicant might decide not to "spend" one of his 12 or so applications on a college where his test scores are closer to or lower than the 25th percentile score -- especially if he has no other particularly appealing trait for colleges (legacy, relative of donor, athlete, not white or Asian).
3. You're right that there is no publicly available data that tell us the marginal effect of not submitting test scores. The best data we have come from the Students for Fair Admission lawsuits against Harvard and UNC, which cover a period before test-optional policies were common. Litigation pried it out of them.
4. Even ostensibly test-optional colleges let in score submitters at higher rates. Why?
a. One possible reason is that intelligence is the lurking variable here. Smarter kids get higher test scores, which they are then proud to send in, AND exhibit other traits that admission officers like (e.g. lots of AP classes).
b. Related: your intuition that colleges assume low test scores for non-score submitters is probably right.
c. An SAT score is like a credit rating from S&P. Non-score submitters are riskier bets (e.g. "will this kid wash out and pull down our 4-yr graduation rate?"), because colleges know less about them. GPAs reflect both grade inflation and ability, whereas standardized tests are standardized.
5. Test-optional policies probably do enable colleges to let in some low-scoring applicants who belong to the favored groups listed above without those scores pulling down the average test score for the US News ranking.
Please let me know if there's anything else I can do for your family!