Interpreting the New Concordance for SAT and ACT

It depends on your test scores thus far and on your target colleges. In most cases, your fall plans – whatever they had been – should probably remain in place, based on factors you had considered prior to the new concordance. You have not moved in terms of relative standing among other test takers of the same test (SAT or ACT), but if you were to concord all scores from a given college’s applicant pool to one scale or the other, you might slightly gain or lose ground. Looking at your score that way may prompt you to consider one more sitting if time allows. Especially if you have only tested once and were not entirely satisfied with your score, you should re-test. Most students are better off re-taking the test they have already taken rather than jumping to a different test. Contact your Compass director to discuss your specific case if you are unsure.

The new concordance hasn’t made either test easier or harder. It hasn’t even made attaining a given score on either test more or less common. It has simply made the highest SAT scores slightly more valuable than they had (erroneously) been in terms of ACT scores. The changes in value are small though, and limited to only part of the scale. Other factors related to test content, structure, and pace could have a greater impact on performance differences. Students are still advised to try a practice version of each test and compare results; just be sure to use the new concordance tables when comparing. We expect the same overall distribution of roughly 50% falling in “Judgment Call” range and about 25% leaning slightly toward one test or the other.

The simple answer is that the SAT scale has more possible scores than the ACT scale. There are 121 values between 400 and 1600. There are 36 ACT Composite scores. It’s not surprising, then, that each ACT score corresponds to 3 to 4 SAT scores.

While there is always a single concordant ACT score for a given SAT score, an ACT score is concordant with a range of SAT scores — a result of the scoring differences mentioned above. In some situations, it is easier to have a point-to-point concordance. For example, a 36 ACT is concordant with an SAT range of 1570-1600 or, as a single score, a 1590.

Either, neither, and both. ACT and College Board do not provide guidance to colleges on the subject. Concordances are created to allow students and colleges to view things in either direction.

In holistic assessment of a student’s testing record, colleges are often looking at more than just an overall score. So even if a college determines your “best score” via a concordance, it may still look at your English, Reading, Math, and Science scores.

Colleges can also develop their own concordances or decide that admission offices should weigh scores without the need of a concordance.

ACT and College Board have said that the new concordances should be the definitive source starting with students applying for terms after fall 2018, and we expect that most colleges will follow this advice. ACT, College Board, and the NCAA have worked hard to produce a definitive concordance, so there is little reason to stick with something outdated. ACT and College Board cannot enforce this practice, however. There have always been some colleges that claim to go concordance-free or that develop concordances based on their own applicant population, but these are the exception.

We won’t know the answer to this right away. While we expect most colleges to incorporate the concordance immediately for admission, there is no assurance that cut scores will be changed for the upcoming applicant class.

A number of scholarship programs also use the concordances to create cutoffs. For example, the University of Alabama offers excellent academic scholarship opportunities and uses the 2016 concordance in an effort to be fair to ACT takers and SAT takers. A Presidential Scholar needs an ACT score of 33 or higher or an SAT score of 1490 or higher. The new concordance, if implemented, would expand the SAT range to 1450 or higher or would involve moving the ACT score higher. Scholarship programs with similar cutoffs exist at dozens of universities and will need to be updated. The timeline for such updates is uncertain.

The NCAA played a critical role in developing the new concordance and will almost certainly be using the new study. The exact how and when is still to be determined. Traditionally, the NCAA has used a sum of the four ACT test scores (4-144) rather than the Composite, so its table is not identical to the 2016 concordance. Given the movement in the concordance in the under 900 range of the scale, SAT takers may actually need higher scores after adoption of the new concordance. For example, an SAT taker is currently a Full Qualifier with a GPA of 2.5 and an SAT score of 900. The new concordance implies that the student would now need to achieve a 920. Given the possibility of moving the goalposts on students who had already been thought to have qualified, it will be interesting to see exactly how the NCAA implements the new concordance. We will update this information as soon as it becomes available.

College Board and ACT specifically instruct colleges not to superscore across exam types — sometimes called super-duper scoring — but colleges are not required to follow this advice. Only a handful of colleges currently super-duper score. We are not yet aware of any colleges changing policies at this point. The subject level concordances are more commonly used for placement purposes or for institutional research.

No. Colleges report SAT and ACT scores separately. The new concordance does not raise or lower either set of scores.

This is only answerable if some simplifying — and perhaps unrealistic — assumptions are made. If the concordance change were the only variable — i.e. no change in student or college behavior — then the revised concordance would likely lower average SAT scores slightly at the most competitive colleges and put upward pressure on ACT scores. The reason the assumptions are unrealistic is that students and colleges have been undergoing a rapid adjustment to the new SAT. The concordance is only one of many moving pieces. At colleges admitting students primarily across the 900-1300 SAT range, it is unlikely that the concordance change will impact overall class composition. The new concordance itself may also affect student testing behavior.

Keep in mind that a student’s relative standing within a test type is unchanged. A student with a 90th percentile score on the SAT will still have a 90th percentile score on the SAT and will be in the exact same competitive environment with fellow SAT takers pre- or post- new concordance. It is only across test types where the balance shifts slightly.

Developing a solid concordance requires lots of data from a representative sample of students taking both exams. Since the new SAT only became operational after the old SAT had been sunset, College Board had to use a more limited pool of approximately 5,000 pilot testers to create the old SAT to new SAT concordance. This provided enough information in the center range — where most students score — but fell short at the tail ends of the curve.

Had College Board and ACT collaborated prior to the release of the new SAT and immediately started putting operational data to use, it’s possible that a concordance could have been released last summer. ACT, of course, had no interest in smoothing the way for the new SAT. Only pressure from colleges and the involvement of the NCAA — the neutral data intermediary — allowed the new concordance study to move forward.

In the long run, it’s best that the concordance was done right. The class of 2018 was the first full cohort to operate (almost) completely in the new landscape. Students were not split between the old SAT and new SAT as they were in the class of 2017 and balance had largely been restored to the ACT/SAT universe.

College Board will not be revising its old SAT to new SAT concordance. Counselors wishing to compare scores should continue to use the 2016 concordance.

There are several reasons that would explain why the 2018 SAT/ACT concordance differs from the 2016 concordance. The original derived concordance depended on 3 assumptions:
The 2008 ACT/SAT concordance developed by ACT and College Board was still valid.
The 2016 old SAT to new SAT concordance developed by College Board from pilot data in 2014-2015 was accurate and applicable to the full testing population.
Chaining together the concordances in (1) and (2) could produce a concordance between ACT and the new SAT.

There is reason to believe that the first assumption was flawed. Both testing organizations saw large increases in test-taker numbers over the last decade, but the growth has not been constant across the tests or across the scale ranges. The number of students taking the ACT grew by 56% between 2007 and 2017, for example, but the number of perfect scorers increased ninefold. Disproportionate growth was also seen at the low end of the scale, as more states required the ACT for all students. This unequal growth did not guarantee a drift in the concordance, but it made it far more likely. In other words, an updated ACT/SAT concordance was probably overdue.

Developing concordances is a lengthy and expensive process. And, quite honestly, College Board and ACT don’t like working together. Concordances are usually updated only when there is a significant content change on one of the exams. The recent situation was unique because of the fact that the SAT was overhauled entirely. The 2016 concordance was really just a spare tire that needed to be replaced as soon as a full concordance could be completed. We shouldn’t have to worry about this for another decade.

It’s complicated. The new concordance adjusts the way that SAT and ACT scores are compared. It does not in any way change a student’s or class’ scores. In theory, SAT scoring has been stable since the new SAT was first offered in March 2016. Forms are carefully equated to one another. The old SAT to new SAT change did, of course, shift scores. High schools can use the College Board’s concordance table in an effort to restate a student’s old SAT scores for comparison purposes.

In analyzing score changes at high schools, we have found that the likely suspects involve changing student behavior (self-selection bias) and profiles. The mix of SAT and ACT takers at a school plays a large role, and that mix has been a moving target. We have seen high schools that went from 90% of its students taking the SAT in the class of 2016 to 10% taking it in the class of 2017 to 50% taking it in the class of 2018.

At many schools, students in the class of 2017 avoided the new SAT by finishing testing early with the old SAT or moving to the ACT. The decisions were not spread uniformly and had unintended consequences. For example, the highest-performing students — ones who often finish testing early — were more likely to opt for the old SAT.

More and more students are at least “sampling” each exam. A student might take an SAT in fall of junior year but then decide on an ACT path. That student’s SAT score would be lower than had they continued on an SAT path. Many high schools compute their average SAT scores by simply averaging their students’ performances (or best performances). This can have the effect of making average SAT scores — and ACT scores — look lower.

The natural fluctuation of class size and composition can easily be mistaken for test-related score changes.

The process of concordance is different than that of equating — determining comparable scores on different forms of the same exam. First, equating different SAT forms avoids the biggest shortcoming of concordances — the fact that they involve different educational constructs. The other major problem with the 2016 concordance is that it had to be based on pilot study data. New test forms, on the other hand, can be developed by linking to operational exams. A given SAT date may see 100,000 to 500,000 students. That provides a rich source of data. Compass has been critical about the mishaps that occurred in the new SAT rollout, but we believe that keeping scores consistent is a core skill maintained by College Board.

Not really. The key is that students and (most) colleges were working off of the same information. A student interpreting his 27 ACT score as higher than his 1270 SAT score (concordant to a 26) would have been correctly interpreting the concordance. And an admission office would have interpreted things the same way.

The new concordance does not change the decision-making process, but it is critical that students understand which concordance is being used by a college. Websites or old forum posts can sometimes provide misleading information. Please contact Compass if you need any assistance in getting your students new materials.

The PSAT and PreACT use common scales with the SAT and ACT, respectively, to allow students and counselors to directly compare scores. For example, a 1000 means the same thing on the PSAT as it does on the SAT. So while the PSAT and PreACT were not part of the concordance study, the same set of tools can be used to gauge a student’s relative standing across test types.

While PSAT and PreACT scores can rely on the new concordance tables, you should factor in the datedness of scores from 9th and 10th grade. What may appear as performance disparity across tests may have less to do with test differences and concordance updates and more to do with improving skills over the passage of time.

We hope so.