If you are a prospective college student who’s getting ready to apply to college in the next couple of years and you’re wondering what’s going on in the test-optional debate, we have some new evidence that has come out, and there was also a recent New York Times op-ed about it. I’m going to talk about it in this blog and also what that means for you if you’re considering going test-optional in the coming years.
So, let’s talk about this new data. A recent New York Times op-ed featured some of this data, but it dropped in January of 2024. There was a recent study from an organization called Opportunity Insights out of Harvard, and they were looking at the performance of test-optional policies at Ivy Plus universities, which is all the Ivies plus Stanford, MIT, Duke, and UChicago. Here’s what they found.
Number one: at elite colleges, test scores were far more predictive than GPA and were highly predictive of freshman GPA. They found that 0.43 points of your GPA correlated with your SAT or ACT score. So, a perfect SAT score correlates with 0.43 more points in your GPA than a 1200 on the SAT. And it is true that only 5% or something of the students who got into these universities had a score as low as 1200 to 1400. So, there’s not a lot of data points in the lower ranges, but they found that test scores were the best predictor of freshman performance at Ivy Plus schools.
Another really interesting thing that they found, and this to me is probably the most interesting takeaway for my students who are trying to strategize on test-optional, is that students who were test-optional and did not submit scores performed about as well as those who submitted scores and were about a 1307. And 1307 is below the 25th percentile at all these schools in terms of their median test scores. If only 5% of students at Brown, for example, were getting between a 1200 and a 1399 on their SAT, you can tell that this is an infinitesimal percentage of students that were getting in with test-optional policies.
Actually, this is kind of interesting. Let’s look at my best bets for test-optional. If you guys haven’t seen it, I have a really interesting chart on our Best Bets for Test Optional video. If you haven’t watched it, you should watch it because it’s interesting to see all this data. I’m just doing a little quick case study on Brown. For Brown, the sum of SAT-ACT submissions last year was about 81%. I think there was a student survey that said it was even a little bit lower than that, that it was around 76% or 75% of students submitted test scores. That means 25% of students didn’t submit test scores, and they were performing at the level that represented 5% of those with test scores. My guess is this is going to affect admissions policies given this kind of performance data, which is new. We didn’t have the data before because colleges were experimenting, and some of the results of those experiments are coming back, and that’s going to change how colleges use data moving forward.
The second finding was that high school GPA was basically not predictive at these top universities. Now this being said, I will say that in a test-optional world, there are probably more variances in test scores among students, I would imagine, than necessarily GPA because of grade inflation. Something like 44% of American high school students have an A average or above, but they found that there was only a 0.1 difference in GPA freshman year at these elite colleges between students who had a perfect GPA at their high school or students who had an unweighted 3.2, which is a super minimal difference in GPA.
So, what does this mean? I mean, first of all, let’s look at the common data set. Only 5% of Brown students had a 1200 to 1400 SAT score, whereas 95% of them who gave scores had a 1400 to a 1600. Okay, so 100% of students at Brown were in the top half of their high school, 97% were in the top quartile, and 93% were in the top 10th. So, a really small sample size in terms of the number of students who are going to have a 3.2, because given what I told you, 44% or about half of students have an A average—which is maybe an unweighted 3.5—so half of American students have like a 3.5, 3.6, or above. A 3.2 is going to be in the bottom half. Brown didn’t have anybody, even in that segment. And so I will admit that there’s going to be a small sample size, but in any case, you can look at the line on the graph and see that there’s a correlation. So, the SAT and ACT scores of admitted students are more predictive of success than their GPA.
The third finding is that low-income students with the same test scores performed about as well as high-income students with the same test scores. I know a lot of the narrative likes to say that test scores are only an indication of wealth, and they actually don’t indicate that much intelligence, so when we use test scores, we’re really just letting in the rich kids who had lots of prep to fake their smarts. Well, the fact that it correlates with GPA kind of discounts that. And the fact that if poor kids can get the same scores as the rich kids, they perform as well in college, and if they get lower scores, they perform about as well as the rich kids with the lower scores, kind of blows that out of the water a little bit. Whether you’re rich or poor, if you can get a great score, it can potentially mean something for your academic performance at these schools.
So what does that mean if you are a parent or a student looking into the jaws of college admissions at these elite institutions? Number one: it means you better darn well take the SAT and the ACT because we might see some schools return to either a test-required or test-recommended policy where they want to see your test score because suddenly it’s correlating much better than GPA with your performance at these schools. You need to be ready for changing policies because, given the state of this data that just dropped in January 2024, you need to be ready for it. Likewise, I am saying at this point to my students that if you’re applying test-optional to Harvard, Stanford, Princeton, or Yale, where they’re always at the top, they’re probably going to assume you’re going to clock in somewhere around a 1307. So, if you’ve got a 1400 or a 1450 and you think it’s below the 25th percentile at Harvard, I think you’re still better off submitting because otherwise they’re going to look at you and think that on average, you’re probably going to perform as well as people with a 1300. And I would rather submit that 1450 and cross my fingers. Obviously, if your score is not high, your chances of getting in are lower, period. But if you’ve got something that’s in between or a 1400 plus, I’m probably going to say from here on out, submit at the very top colleges, at least those four HYPS, as we like to call it.
Number two: don’t sweat your GPA, especially if you go to a super competitive school. It looks like GPA doesn’t matter that much. So, if you get a couple of B’s, maybe you just have some mean teachers.
Number three, though many colleges will likely continue test-optional policies officially, they are likely to admit fewer and fewer candidates without scores given this data that’s come out. So, I expect an uptick in the consideration of scores and an uptick in terms of colleges expecting to see scores. Even if colleges say they’re test-optional, if you are test-optional, your chances of getting in are probably going down given this data drop. I think colleges are going to be looking for scores. Test scores can matter; they are predictive, so you might want to get on that.
I hope you guys enjoyed this blog and that it was helpful!