“Smoking-Gun” Data on North Carolina ACT Scores

<ɫTV>Students who submitted scores during the test-optional period had better freshman grades than those who didn’t.

Earlier this year, the UNC Board of Governors approved a new system-wide admissions policy requiring standardized tests only for students whose high-school GPAs are less than 2.8. This comes after years of testing waivers that began in 2020 as a response to the Covid-19 pandemic and the limited availability of tests.

The new policy provides an important additional metric for many schools in the UNC System. But it is meaningless at the two most competitive public institutions. I wrote at the time:

[T]he new policy would make the System’s most competitive schools—UNC-Chapel Hill and [NC State]—effectively test optional. This would make it harder for admissions officers to make distinctions between the thousands of students whose GPAs qualify them for admission. At UNC-Chapel Hill, for example, 95 percent of freshmen admitted in fall 2022 had a GPA of 4.0 or higher. None had a GPA of less than 2.99. Admissions officers would be forced to rely on more subjective and less reliable admission criteria such as personal essays and letters of recommendation.

New data confirm that the policy is indeed inadequate.

During the pandemic-era test-optional period, many students who attended public high schools in North Carolina still took the ACT during their junior years, as required by North Carolina law. This set up an interesting natural experiment, since the North Carolina Department of Public Instruction collected these ACT scores from high schools and shared them with the UNC System. Therefore, we know the ACT scores of all public-university students who applied to UNC-Chapel Hill and NC State during the test-waiver period, regardless of whether they submitted those scores for use in the admissions process.

Almost certainly due to their lower academic qualifications, non-submitters fell immediately behind their peers.The Martin Center obtained these comparative data from the UNC System via a public-records request. They are revealing. At both NC State and UNC-Chapel Hill, the test scores of students who chose to submit them as part of the admissions process were significantly higher than those of non-submitters.

And these discrepancies persist among admitted students. In 2022 (the year for which the best data exist), among admitted UNC-Chapel Hill freshmen, test submitters had a median ACT score of 29. Non-submitters had a median ACT score of 25. At NC State, the respective scores in the same year were 27 and 22. These are meaningful differences.

The data below are for Fall 2022. Complete data (Fall 2021-Fall 2023) for UNC-Chapel Hill are available . For NC State, .


Almost certainly due to their lower academic qualifications, non-submitters fell behind their peers during their first semester on campus. This is what we expected based on Mismatch Theory: that students with lower standardized-test scores will receive lower grades once admitted, fall behind, become frustrated, and sometimes switch to easier majors or drop out.

Without a policy change, UNC-Chapel Hill and NC State will continue to admit less-qualified applicants. Standardized testing is a necessary tool to judge academic preparedness and predict academic success, especially at competitive universities. The data now further prove it.

Jenna A. Robinson is president of the James G. Martin Center for Academic Renewal.