Can test scores and grade-point averages (GPA) predict success in college? Huy Le, professor of management at the UTSA College of Business, has built a career around this question. Most recently his work on college retention was published in Education Assessment and featured in Forbes.
“My interest is in individual difference,” said Le who has consulted with ACT, Educational Testing Service and the College Board. “How are individuals different in terms of abilities, personality and interests in general? And how we can apply that knowledge about individual differences to improve organizational effectiveness and education.”
Applying this knowledge piqued Le’s interest in standardized testing. He and his coauthors first published in this area in 2015. Using data from 50 institutions and more than 180,000 undergraduate students, they looked at different predictors such as standardized test scores, high school GPA and socioeconomic status to see how they relate to college performance.
“In the first study we looked at each of the predictors separately,” said Le, who has taught at UTSA since 2014. “In the second study we used the findings from the first study and combined the factors together to predict the probability that students will stay in college after the first year.”
In the initial study, ACT test scores and high school GPA were highly-correlated to the first year college GPA. The findings showed that socioeconomic status was a weak predictor. Individually these results were valid, but when combined, test scores and GPAs were more accurate.
National data on withdrawal rates from four-year universities shows that the highest dropout rates are seen between the first and second academic year. Le’s research found that first-year academic success may serve as the central vehicle for retention efforts.
“One surprise that we saw was that all of the factors are important to student retention, but their effects only influence the first-year GPA,” he said. “In other words, first-year GPA is the most important predictor of retention.”
During the pandemic most colleges have waived standardized testing requirements or made them optional. But Le believes they’ll eventually go back to utilizing standardized testing because there are not other methods for them to use.
“There is a lot of debate about standardized testing, and if it is helpful,” Le said. “There are definitely limitations. But these limitations are better than not using this type of data at all. Using scores, plus high school GPA can help improve the predictions of outcomes and identify students who will be most likely at risk. Based on this data, schools can identify these students and help them much sooner.”
Looking to extend upon this research, Le and his colleagues plan to conduct similar studies with graduate students using testing such as the GRE and GMAT. “We expect we’d find something similar in results, but we need to get the data and results,” he concluded.