New SAT Easier than ACT?
Georgetown recently announced the admission details for the class of 2021. In addition to reaching a record low acceptance rate of 15.4% and the largest number of applicants in its history, the university also reported the average SAT and ACT scores for accepted students. For the first time since the redesigned SAT was implemented, we now have real data indicating how colleges are using the redesigned SAT results in admissions decisions, and there is an interesting and concerning difference in SAT and ACT percentiles that might raise doubts as to how colleges are comparing the two tests.
|Test||25th percentile||75th percentile|
|Test||25th percentile||75th percentile|
|SAT EBRW||690 (92nd)||770 (99th)|
|SAT Math||680 (89th)||770 (98th)|
|ACT||31 (96th)||34 (99th)|
At the upper end of the range, there isn’t much of a surprise; the upper range for both tests maxed out at around the 99th percentile. In other words, Georgetown is a competitive enough school that the top 25 percent of the applicant pool are scoring in the 99th percentile on the tests.
At the lower end of the range, we see a concerning discrepancy between the SAT and ACT submitters. It appears that an average student at the bottom of the SAT range was somewhere between the 89th and the 92nd percentile of the applicant pool. By contrast, the average ACT submitter at the bottom of the range was in the 96th percentile. In other words, an SAT submitter needed only to be in the 89th-to-92nd percentile to be considered in “average” range. An ACT submitter, on the other hand, needed to make it up to the 96th percentile to be considered “average.”
This data point suggests that, at least on the SAT/ACT front, the admissions cycle was more forgiving for SAT-submitters. Though Georgetown’s dean of admissions noted that scores on the redesigned SAT were higher than the old test, it appears that the percentiles for those scores were lower. In other words, the score makes a student look more competitive than an ACT-submitter, even though the percentiles indicate he or she is actually less competitive.
Strangely, the notion that the SAT is an easier test was directly bolstered in a College Board press release that presented the results of a student survey on the redesigned SAT. Seven times in the press release, the College Board highlighted student comments that the test felt “easier” than the previous test. While potentially an attractive notion for students and families who aren’t necessarily interested in taking the more difficult test, an easier SAT could lead to potential problems with admissions offices: a higher SAT score since more students are getting more questions correct, but a lower percentile because no matter how well students do collectively, half of the students taking the test must be considered below-average for percentiles to make sense.
If this pattern continues to bear out in other admissions offices – of higher SAT scores than the previous SAT but lower percentiles when compared with the ACT – we can only imagine the confusion that will result in other admissions offices with less-equipped officers, not to mention families who are wondering which test they should take. Should I take the ACT because the College Board conversion table puts me in a higher percentile with that test, or should I take the SAT because it seems to be viewed more favorably at admissions offices? That will be a difficult question to answer until the College Board clarifies the relationship between its scaled scores and percentiles.
If more colleges report percentile discrepancies like those of Georgetown, the burden lies on the College Board to justify the score validity of the new test, to share with colleges and families percentile tables based on real student data, and to educate admissions officers on exactly how to interpret redesigned SAT scores. This would also necessitate collaboration with the ACT to ensure that both tests are accurately indicating a student’s college-readiness and not just trying to sell the higher-education world on the superiority of its test.