Why the ACT-SAT Concordance Isn’t Totally Boring

 

ACT and the College Board have released the concordance table for the ACT and redesigned SAT.  There’s been a correction that leans in favor of the SAT at the high end of the scores and against the SAT at the low end. The concordance is also a powerful reminder that the tests are being misused by universities and colleges.

Last Thursday, in the midst of the news that the University of Chicago had gone test optional, ACT and the College Board released their new concordance table, which allows students and schools to compare an ACT score to an SAT score.  Download it here.

As we expected after looking at how colleges dealt with the redesigned SAT in year one–and seeing that average SAT scores did not go up as much as the concordance would lead one to expect–the new table is more forgiving to the SAT than the previous one. That is sort of ironic since the previous concordance was designed without ACT’s collaboration, a fact that ACT did not let slide.

The equivalent ACT score for SAT scores above 1250 increased by 1 point across most of the range, meaning that a high SAT score now looks stronger against the ACT than it did last year.  For example, a 1400 on old concordance was equated with a 30 on the ACT; on the new test a 1400 is equivalent to a 31.  The middle range of SAT scores, between 950 and 1250, where almost half of all students score, largely stayed the same. At the low end, from 590 to 940, the SAT is a point or even 2 points worse against the ACT.

Scores marked with an asterisk in the table below indicate the SAT score the concordance identified as best for users who want to concord an ACT score to a single SAT score point.

SATACTChangeSATACTChange
16003601090210
*159036+1*1080210
158036+11070210
157036+11060210
15603501050200
155035+1*1040200
*154035+11030200
153035+1102019-1
1520340*1010190
151034+11000190
*150034+1990190
149034+198018-1
148033+1*970180
147033+1960180
*146033+195017-1
145033+194017-1
144032+1*930170
*143032+1920170
142032+191016-1
141031+190016-1
*140031+1*890160
139031+1880160
138030+187015-1
*137030+186015-1
136030+1*850150
1350290840150
*134029+1830150
133029+182014-1
132028081014-1
*1310280*800140
130028+1790140
1290270780140
*128027077013-1
127027+1*76013-1
126027+1750130
1250260740130
*1240260730130
123026+172012-1
1220250*710120
*1210250700120
1200250690120
119024068011-1
*1180240*67011-1
117024066011-1
116024065011-1
115023064010-2
*1140230*63010-2
113023062010-2
11202206109-2
*11102206009-2
1100220*5909-2

The new concordance does something that earlier concordances (1994, 1999, 2009) did not:  it provides tables for the math sections and for the reading and grammar sections.  The 2009 concordance did concord ACT’s combined English/Writing score and the previous SAT’s Writing score. The reason the test makers were not able to create a concordance between the other sections in the past is probably that the scores among test-takers who took both did not correlate closely enough on the individual sections to compare them.

It is no big surprise that ACT and College Board did not bother to equate the essay sections.

Some people might wonder why we need concordances at all.  Couldn’t we just compare the percentiles to each other? Isn’t getting a 96th percentile score on the ACT Composite, which is a 31, the same as getting a 96th percentile score on the SAT, which is a 1420? According to the new concordance, no. A 31 on the ACT concords to a 1400 and 1420 on the SAT concords to a 32.

Percentile comparisons are not as accurate as comparing the scores of students who took both tests because the pools of test takers for the ACT and SAT are not the same. The percentiles and pools do overlap, which is what made it possible for ACT and College Board to compare the scores of 589,753 students in the Class of 2017 who took an official SAT and official ACT.  The comparison of actual test takers is more accurate and suggests that, overall, the pool of students taking the SAT is a little stronger than those taking the ACT, which has been true in the past as well.

One good feature of the new concordance is that it provides the option to equate an ACT score to a single SAT score.  This feature is most relevant to the schools that use cutoff scores to award merit aid.  As we discovered in our collaboration with Edmit, many schools use cutoff scores, but there were variations in how they decided to match SAT ranges with precise ACT scores, which provided. ion some cases, advantages to those who took one test over the other.  We hope that schools will use the new concordance to reset their cutoffs.

Better yet, we hope that schools will look at the concordance and decide to drop cutoff scores altogether.  There is a fairly stunning note at the end of the guidance ACT and College Board released, which serves to remind us all that test scores are not nearly as precise as we might like to think they are.

When using the SAT Total and ACT Composite concordance table to estimate a student’s proximal ACT Composite score from their SAT Total score, the estimates in the table have a standard error of approximately ± 2.26 (2) ACT Composite score points on its 1–36 point scale. When using this table to estimate a student’s proximal SAT Total score from their ACT Composite score, the estimates have a standard error of approximately ± 79.57 (80) SAT Total score points on its 400–1600 point scale. (6)
In other words, the concordance has a large margin of error. if you have a student with a 25 on the ACT, the concordance tells you that’s equivalent to a 1210 on the SAT, but with the standard error , that 25 equates to a much less precise range of about 1130-1290.  Another way to think about that is that an 1130 is the 64th percentile and a 1290 is the 87th percentile, so the 25 on the ACT concords to a range covering 25% of the SAT curve.  Colleges and universities surely do not want to decide who to award scholarships to based on something so deeply imprecise.

 

Please follow and like us:

Add a Comment

Your email address will not be published. Required fields are marked *

error: Content is protected !!