Quirk of the ACT Test — Rounding to a Whole Number — Offers Rare Isolated Glimpse of the Value of Scores

Tracking the impact of a small test-score difference on college attendance and later life

The household drama around college admissions is not only due to the stakes involved — where one studies has a large impact on career prospects, lifelong income and, yes, it seems, self-worth — but also because the process itself can seem so maddeningly opaque. Funnel in your high school senior’s test scores, grades, a sparkling essay and some impressive letters of recommendation and hope it all somehow grabs the attention of a college admissions officer.

Little wonder that thriving industries have popped up to help burnish submissions and decode the leanings of universities. If tutors and test-prep classes, which cost money, improve scores, what does an extra point on its 36 point scale get you?

A quirk of the ACT — that it combines four subtest scores and then rounds to the nearest full number — allows UCLA Anderson’s Kareem Haggag, Brigham Young’s Emily Leslie, University of Chicago’s Devin Pope and University of Maryland’s Nolan Pope, in a paper published in The Journal of Human Resources, to offer an estimate. Nearly identical scores — one rounded down to 24 from 24.4, for example, and one rounded up to 25 from 24.6 — give the rounded-up student a 0.44 percentage point higher likelihood of attending a four-year college.

Standardized testing, with the stated intention of putting meritocracy ahead of wealth and connections, spawned a system in which households with the most resources could further improve their odds with tutors and test prep. The authors note that those same households are likely to submit college essays (some with the help of consulting editors) and letters of recommendation of perceived high quality.

Is Test Prep the Answer?

By isolating the role of test scores across the more than 1 million ACT takers annually, the researchers provide valuable data to families weighing investments in test result improvements. And because other research has shown that test prep and tutors generally help improve scores, this research helps show how even small improvements can produce a meaningful result.

The “random” distinction between the rounded-down and rounded-up test takers, the authors suggest, might help in the wider, ongoing evaluation of inequities in allocating scarce seats in college programs.

The ACT is effectively four mini tests covering English, math, reading and science. (The science component became optional in April 2025). Each section is scored from 1-36 and then a single Composite ACT score is the average of those four scores.

“The data used for this project were provided by ACT, Inc. and include the test scores and other student characteristics for all ACT takers who graduated from high school in 2011 and 2012. These data were merged at the individual level with data from the National Student Clearing House that indicates the college, if any, that each student attended. Importantly, the data contain the four ACT subscores, which allow us to create an unrounded score for each test taken.” 

The researchers also layered in datasets that report graduation rates for ACT takers and demographics on parents of ACT takers, as well as the future average earnings of graduates at age 35. 

Because students who receive a rounded-up score are less likely to retake, the raw 0.44 percentage point increase in four-year enrollment may understate the impact of the score itself. After adjusting for retakes, the authors infer that a one point increase in a student’s final (maximum) ACT composite raises the probability of enrolling in a four-year college by about  0.67 percentage points.

Earlier research found that retaking the SAT test led to higher scores and a higher likelihood of attending a four-year college.

While the effect of ACT rounding is subtle, that it exists at all is another data point in the limitations of standardized testing. Moreover, the researchers find that the impact of a rounded-up score feeds a potentially consequential shift in school choice: Students are more likely to opt for a four-year school than community college. And students who benefited from rounding up were found to apply to more selective schools.

The Effect of the Round-Up on Graduation

Benefiting from rounding up doesn’t, however, presage higher college graduation rates.

“The researchers estimate that students who benefited from rounding up attended schools at which graduates earned slightly higher annual incomes at age 35. This seems to be a function of having attended more selective schools, which was aided by the fact that students whose score was rounded up applied to more selective schools.”

While nearly 6 in 10 high school graduates took the ACT as recently as 2016, fewer than 40% took the ACT in 2023. Still, that translates to nearly 1.4 million students whose college fate relies in part on the ACT. 

The financial inequity in the test-taking ecosystem, along with concerns of whether the actual test structure captures valuable insights on student capability, has led some schools to move to a “test optional” admissions system. Legal pressure pushed the University of California system to stop using SAT and ACT scores in its admission process four years ago.

There is no uniform data on how the ACT Composite score is used in the admissions process for schools that still rely on standardized testing. But it is acknowledged to be a popular vetting mechanism along with other factors, including grade point average  and the raw underlying subscores for each component of the ACT. Moreover, plenty of schools require a minimum Composite ACT score to be eligible for specific merit-based scholarships and grants. Given the coarse rounding scoring mechanism, that means plenty of students each year are at the mercy of whether their score is rounded up.