You Have Your March SAT Scores – What Now?
The much-anticipated March SAT scores were released earlier this week. So, what can you make of them? Is your 1200 on the new test better than a 1200 (Critical Reading + Math) on the old, and if so, how much better?
True to its promise, the College Board also recently released concordance tables to compare scores on the old SAT with those on the new SAT. These tables, the missing key to deciphering redesigned SAT scores, are much more helpful than the preliminary PSAT tables, as they allow students to compare their March redesigned SAT results with those of the now-retired SAT. More importantly, the College Board has provided a means by which to compare the results of the new SAT and the ACT, a much-needed tool that allows families to decide which test better suits their child.
However, there are several things to keep in mind when using this newly released resource:
First, the College Board has stated (p. 7) that the scoring system for the new test is different from that of the old test. It’s a bit deceptive that both tests are scored on a 200-800 range per section; a 520 on the old test is not equivalent to a 520 on the new test (hence the need for a concordance table). According to College Board, a 1400 out of 1600 on the old test (considering only the Reading and Math sections) equates to a 1470 out of 1600 on the new test. Similarly, while the mean score on the old SAT was around 1000 out of 1600, the mean on the newly released test is 1080. Notice that this means a student may feel that she did better on the new test when in truth she performed approximately the same (or even worse).
Second, the SAT/ACT conversion table that the College Board provided does not have the same degree of accuracy as the table they released in 2006. The College Board explains how its score comparison tool compares the two tests: first it converts the new test to the old test, then it uses the 2006 SAT/ACT table. Helpful? Immensely. 100% accurate? Not exactly. You’ll still want to use your best judgment when deciding between the two tests, especially if the scores are close to one another.
The 2016 concordance table differs significantly from the previous one for another reason. The 2006 SAT/ACT table was the result of a collaboration between the College Board and ACT, Inc., and it used real student data from the class of 2006. The 2016 table was created by the College Board alone, and Marten Roorda, CEO of ACT, Inc., is crying foul. Among the criticisms Roorda launches at College Board are the fact that this new table is based on a single administration of the test rather than a more robust dataset and that it relies on chain linking (concording first to the old SAT and then to the ACT). Roorda asserts that any accurate table comparing SAT and ACT results would require at least a year’s worth of actual student data and collaboration among College Board, ACT, Inc., and independent statistical groups. (UPDATE 6/14/2018: ACT, Inc and the College Board have released updated concordance numbers. The newest numbers tell students who score above a 1250 on the SAT that their equivalent ACT score is about 1 point higher than was communicated in 2016’s tables.) . (UPDATE 5/16/2016: The College Board later corrected the ACT that it did not use the March data in creating the concordance tables, but rather two concordance studies in December 2014 and 2015. The company also stated that it had reached out to the ACT to update the tables, with no response.) Regardless of which company wins this statistical argument, we have another reason to use the SAT/ACT concordance tables with caution, to be sure.
Finally, we’re still waiting on percentile charts for the new SAT based off of real student data. On each student’s score report, a “nationally representative sample percentile” and an “SAT user percentile – national” are provided, but those are also based off of research data, not actual student data. As we discussed when the PSAT scores were released earlier this year, there seem to be real problems with both of these so-called percentiles. We will have to wait several years before the College Board releases percentile charts grounded in actual student data. Until then, we have to trust the College Board that a 1080 is the new 50th percentile.
For now, despite their problems, the College Board’s concordance tables are the best available way to make decisions about whether students should prepare for the SAT or ACT. To make it easier to access the information in these tables, we created the Applerouth score converter. You can input PSAT, old SAT, redesigned SAT, or ACT scores and see what College Board says is the equivalent score on the other tests. If there’s a large gap (more than 2 ACT points) between your ACT and SAT scores, focus more of your studying energy on the test you’re scoring better on.
College Board took a step forward this week in finalizing the rollout of its new SAT, but there’s more work to do. This transition from old SAT to redesigned SAT will require parents, students, and counselors to adjust their conventional thinking about scores. Admissions officers around the country will need to educate their staff that “1080 is the new 1000” and “1450 is the new 1400.” And we will all need to continue to watch and wait to see how the score comparison tables shake out once more data rolls in.
Footnote: For this year, while so much uncertainty is in the air regarding the SAT and ACT tests, it might be a good insurance policy to take both the SAT and ACT. You can still prep for one test, but having two test scores will ensure that, whatever the College Board and ACT decide on concordance, you’re will have locked in your best score on either test.