Superscoring standardised admission tests will supersize advantage

On US students’ college-prep checklist, nothing looms larger than college admission tests. That’s why recently announced changes to the ACT might seem like a windfall to students who are concerned that a single test score could determine their whole futures.

Starting next fall, students will be able to retake any of the ACT’s five sections – mathematics, science, reading, English, and the optional writing section – by sitting for them one at a time, rather than all at once again. Students will then earn a “superscore”, calculated by combining their highest scores on each section.

Sounds like a great idea, right? Sure – if you’re a student with the time and money to supersize your test-taking, or if you are a college chasing prestige by boosting the average test score of your admitted class. Those without the resources for tutors and repeated testing fees will just fall further behind, as their competitors pull away from the pack.

This will just widen the gaps in class privilege that are already baked into the tests. And because socio-economic status correlates highly with race, the racial gaps will widen, too.

We’ve long known that the ACT and its competitor, the SAT, are poor predictors of how a student will do in college. For example, the SAT predicts only 16 per cent of first-year GPA on its own and only 23 per cent in combination with high-school GPA, according to a 2010 analysis of data from the University of California. And at the Georgetown University Center on Education and the Workforce, we’ve found that admission tests also tell us little about students’ chances of graduation: at selective colleges, students who score above a 1200 on the SAT (or equivalent ACT score) graduate at a rate of 85 per cent, compared with a slightly lower rate of 79 per cent for those who score between 1000 and 1099. Students generally need at least an 1150 SAT score to get into the top 500 US colleges.

What the tests seem to reliably predict is not students’ future successes but their background characteristics, such as socio-economic status and race – factors that are tied to systemic inequality. In fact, research has shown that the SAT in effect functions as a proxy for these background characteristics: affluent, white and Asian students make up the majority of high scorers.

In other words, standardised tests give a scientific sheen to the exclusion of low-income, black and Latino students while claiming to sort applicants by merit.

In our forthcoming book, The Merit Myth: How Our Colleges Favor the Rich and Divide America, my co-authors and I describe how admission tests that were originally designed to lift talent out of obscurity by measuring qualities such as IQ, aptitude and achievement became a means of perpetuating racial and class disparities.

When college attendance skyrocketed in the aftermath of the Second World War, the tests seemed like a necessary shortcut to identify students with the most promise. But then higher education’s student bodies diversified, and it became clear that the tests don’t measure aptitude or achievement at all. Put on the defensive, the tests’ architects have failed to keep up. The ACT changes are just the latest of the “improvements” they have put in place over the years, which have done little more than create new loopholes for the already advantaged. 

The test companies are well aware that they still have a problem with fairness, and they know it is a threat to their business model. That’s why they continue to try out ways to make the test more equitable, as when the College Board, which runs the SAT, announced earlier this year that it planned to launch an “adversity score” to give socio-economic context to students’ SAT scores. The plan was quickly scrapped in the face of criticism that it was impossible to sum up the individual challenges faced by students in a single number, but the issue it was intended to address remains.

So who stands to benefit from the ACT change? The answer seems to be the ACT organisation itself. The new rules provide a strong incentive for students to take the test repeatedly, and they may give the ACT a competitive edge over the SAT.

But while critics howl at the continued unfairness of the system, those who have relied on standardised tests to maintain their privilege will be loath to give them up. And while standardised tests are overused, they do play a role in holding colleges accountable for their admission decisions.

Research has shown that “test-optional” admissions – an increasingly popular practice that allows students to apply without submitting standardised test scores – allow colleges to boost their average scores because lower scores are less likely to be submitted. But this is more beneficial to institutional prestige than to low-income or minority students.

In the end, there’s no easy solution to the warped meritocracy of admission tests – but the ACT’s latest plan will help only those who don’t need the assistance.

Anthony P. Carnevale is founder and director of the Georgetown University Center on Education and the Workforce and a former vice-president of Educational Testing Service, which develops and administers the SAT.