A Progressive Case for Standardized Testing
How I Learned to Stop Worrying and Live with the ACT
For generations of high school students, the ACT and the SAT have been the be all and end all of college admissions. Some students prep for years before taking the exam, and many high schools even offer specific classes about getting a better score. It’s not just about getting into your top school though, because real money is on the line; the University of Missouri, my home institution, gives out automatic scholarships on the basis of standardized test scores. A score of 20 on the ACT can get you $1000 dollars a year. A 35 can get you $7500. If you get a 36, you get a full ride scholarship, paid room and board, and even a $3500 stipend. The policy is unusually generous but not totally anomalous given the immense value placed on perfect scores. The unfortunate fact is that students are right to worry and care a great deal about standardized tests.
That one such metric has such a profound impact is hard to stomach, and in high school there was little I hated more than the testing grind. However, the issue gains moral weight given how heavily test scores are correlated with family income. It is hard to know exactly what matters most to this correlation, but the list of mechanisms is relatively straightforward. Wealthier students get into better schools whether through private education or because their parents move to a well-regarded school district. Wealthier students are likely to have fewer obligations and more time to study. They can afford more extensive and personalized test prep, and they can also simply afford to take the test more times. Unexpectedly to me, wealthy students also get more time to take the test as they are more likely to get exceptions.1 Given all this perhaps we should applaud the recent moves of schools which have made test scores optional like the University of Chicago and Cornell. Some schools, like the University of California system, have even decided to exclude tests altogether.
Unfortunately, the alternatives are at least as problematic.
Testing the Alternatives
Think again about the mechanisms which make standardized tests unequal. These include access to better secondary schools for the wealthy, time constraints for low-income students, ability to pay for private lessons and materials. All these mechanisms are still in play for other types of admissions criteria like essays or extracurriculars. My argument for standardized tests is not that they are straightforwardly good. Rather, they are good in comparison to a set of alternatives which have serious drawbacks themselves. Within the context of a meritocratic admissions system, standardized tests have a role to play.
The most popular alternative is GPA. GPA obviously suffers from the same problems, plus it introduces the issue of grade inflation which is more prevalent at elite and private secondary schools. This makes sense as the parents of wealthy students are often highly invested in prestigious university placements, having both the means and motive to intervene on behalf of their children. Further, GPA calculations vary significantly from school to school in ways that are hard to control for. This poses a major problem for college admissions systems as well as education researchers. I don’t mean to write off GPA, and there is a robust debate about whether it is more predictive than standardized test scores.
However, I do want to suggest that it is at least as tainted a metric. Additionally, there are many students for whom the rigid discipline of the high school classroom is a very poor environment. There ought to be some way to assess their ability that isn’t linked to the draconian classroom. Eliminating standardized test scores disadvantages this cadre of students.
Qualitative criteria also pose substantial problems. Extracurricular activities are probably worse than standardized tests as they amp up the importance of extra time and money. Essay writing for application prompts can be done with the help of a paid assistant and/or the help of an editor or tutor can have a huge impact. These qualitative metrics also open the application process to more of the implicit bias of application decision makers. A British study found that teachers were more likely to see high income kids as academically competent compared to low income kids with identical school performance. By contrast, a 2016 study found that making standardized testing mandatory increased the number of students in gifted and talented programs. This shouldn’t come as a surprise. Qualitative metrics remain important, and they can certainly be a place for students to acknowledge how and why they deserve a shot despite an unequal hand. However, we should worry about seeing qualitative criteria as the path out.
The Role of Testing
Considering these alternatives reveals a tough truth: all criteria that try to find the most meritorious students will be linked to wealth and income. As long as expected performance remains important for admission this bias will exist; consequently, having a standardized score as part of the ‘criteria cocktail’ is more likely to help than hinder the drive toward equality. The point here is not that admissions should be based only on scores, but that they provide valuable information as part of the admissions process.
This has been born out in real world examples such as MIT which dropped and then reinstated test requirements. At MIT the admissions office found that without using standardized test scores they had a harder time adjudicating between students who would succeed in a college setting. Worse, looking only at GPA and qualitative metrics ended up resulting in more students from private and advantaged schools. According to the Dean of Admissions, “Once we brought the test requirement back, we admitted the most diverse class that we ever had in our history…” This should not be a surprise.
In essence, the SAT and the ACT have a branding problem.They are highly qualified and ubiquitous, making it trivial to run a regression and find that these tests correlate powerfully with income. The ease of that calculation makes the anti-test drum very easy to beat, but it shouldn’t make us less skeptical of other metrics. When the hard qualitative work is done on metrics like essay writing, they have been shown to be at least as bad. One 2021 study states point blank: “Results show that essays have a stronger correlation to reported household income than SAT scores.” Examining for extracurriculars, especially sports performance, is also a major mechanism of inequality. A 2020 study which looked at such preferences found that “removing preferences for athletes and legacies would significantly alter the racial distribution of admitted students.” Again, the implication is clear.
None of this is meant to exonerate the role that testing has played in past discrimination. One such case occurred in the 1950s when the University of Texas at Austin selected a cutoff score which would exclude the majority of black applicants and made it the prerequisite of admission. However, what I would suggest is that discriminators are going to discriminate regardless and the bluntness of the UTA method is at least slightly preferable to the more thorough discrimination that qualitative metrics can make possible. An example of this point comes from Yale in the early 20th century. At that time, Yale used one standardized test for applying students. Those who passed were accepted. But to the chagrin of WASP administrators far more Jewish students were passing than they wanted to admit. Consequently, Yale’s head of admissions introduced interviews, letters of recommendation, and personality testing in 1922, all as mechanisms of exclusion.
Bigger Fish to Fry
Overall, on net, and considering the other options available standardized tests are one of the more fair parts of the current admissions process. This doesn’t mean they are not onerous, but everything about the admission process is onerous. It also doesn’t mean they aren’t used in ridiculous ways, but so much about the admissions process is ridiculous. The problem is the process as a whole, the aims of that process, and the people who run it. Eliminating standardized testing is not the place to start with reform and it may be a step backwards.
If colleges really care about reducing inequality the best thing they can do is eliminate advantages for legacy students. It is telling that elite colleges are so much more willing to boot test requirements than they are to get rid of such admissions. This is the lowest hanging fruit. More audacious colleges could rework the balance of their merit and needs-based scholarships. The most daring schools could even add admissions lotteries that allow automatic admission to students who have experienced substantial disadvantages but meet some set of minimum criteria.
At the end of the day, the problems of inequality in higher ed are not produced by testing or any individual metric. They go much deeper than that.
As long as colleges (and their admission programs) are machines for producing the most successful and wealthy graduates, they will try to select the students who are already bound to succeed. Those students will disproportionately be white and wealthy. The role I see for standardized test scores is in reducing the extent of that disproportion and keeping admissions honest. So long as we have a meritocratic system it should be as transparent as possible. Ultimately though, the progressive energy aimed at standardized tests is better redirected at the very idea of meritocracy. This is the bigger battle, and it involves less concern about elite universities and a lot more attention to the improvement of public universities and community colleges.2 Of course, it goes even deeper, and starts in public schools and disadvantaged communities themselves. Standardized test scores are really, truly not the enemy; there are bigger and better fish to fry. In the words of the old civil rights spiritual, keep your eyes on the prize.
These exceptions are given out to students with conditions like ADHD and dyslexia. It’s a good idea, but one that is unequally applied given different rates of diagnosis based on wealth.
The Ivy League admits about 60,000 students a year. The University of Texas System alone admits more than 110,000.
 | A guest post by
|