News Release

First-semester GPA a better predictor of college success than ACT score

Peer-Reviewed Publication

University of Illinois at Urbana-Champaign, News Bureau

Susan Gershenfeld, University of Illinois at Urbana-Champaign

image: Colleges and universities that want to boost completion rates among underrepresented students may want to pay closer attention to students' first-semester GPAs, which are reliable predictors of whether students will persist to graduate or drop out, suggests new research by Susan Gershenfeld, who conducted the research while earning a doctorate in social work at Illinois. Gershenfeld's co-authors were educational policy studies professor Denice Ward Hood and social work professor Min Zhan, who also was Gershenfeld's thesis adviser. view more 

Credit: Photo provided

CHAMPAIGN, Ill. -- Underrepresented students' first-semester GPA may be a better predictor of whether they'll graduate college than their ACT score or their family's socioeconomic status, a new study found.

Researchers at the University of Illinois tracked the academic achievement and degree status of more than 1,900 U. of I. freshmen across a six-year period, beginning when the students first enrolled at the university in 2005 or 2006. The sample was selected to focus on students who were low-income, attended underresourced high schools and/or were historically underrepresented based on race or geography, and who could have completed an undergraduate program within six years.

The researchers examined the impact of individual characteristics such as race and gender, along with factors such as the academic units and majors freshmen were enrolled in during their first semester on campus.

Of the 69 percent of students who earned diplomas within six years, the researchers found that the composite ACT scores of students who graduated and those who dropped out were nearly identical - 24.5 and 24.1 points, respectively.

Racial minorities, who constituted 93 percent of the sample, graduated at higher rates than did the white students who were low-income or from underrepresented counties within Illinois, according to the paper, published in the Journal of College Student Retention: Research, Theory and Practice.

The freshmen who persisted to graduation had significantly higher first-semester GPAs - 2.84 versus 2.20, respectively - compared with peers who left without earning a degree, according to principal investigator Susan Gershenfeld, who conducted the research while earning a doctorate in social work at the U. of I.

Gershenfeld also is the former director of the university's Illinois Promise program, a scholarship program that covers all of the educational costs - including tuition, books and living expenses - for the most disadvantaged in-state students.

"The goal of the research was to help us better understand why some students are not successful, even when some of their financial barriers to college access are removed," said co-author Denice Ward Hood, a professor of education policy, organization and leadership at Illinois.

In the sample, about 44 percent of the students who enrolled in 2005 and nearly 46 percent of those who enrolled in 2006 received federal Pell Grants.

Most of the students also were eligible for Illinois Promise and/or two other campus initiatives aimed at promoting the access and success of underrepresented students: a merit program that, at the time, provided $1,000 scholarships annually to high-achieving freshmen, and a program that provided support services such as advising and help with study skills - but no scholarships - to students from underrepresented school districts in Illinois.

Freshmen with first-semester GPAs of up to 2.33 were about half as likely to graduate as students who had GPAs in the 3.68 to 4.0 range, the researchers found.

"What this research shows is that students who are above that 2.0 cutoff, but below 2.33, are at significant risk of not graduating. Waiting until a student hits a 2.0 GPA or lower may be too late," Gershenfeld said. "Freshmen with first-semester GPAs of up to 2.33 should be targeted as particularly vulnerable to attrition."

University and federal student aid policies require that students maintain a cumulative 2.0 GPA - a C average - on a 4.0 scale. Students' whose GPAs drop below 2.0 are placed on academic probation and offered services such as advising, mentoring or tutoring.

Universities' approaches to identifying students who may need academic help, based upon their GPAs, and the types of support services offered to these students have changed relatively little over the past three decades, said Ward Hood, who was an academic adviser to at-risk students at another university early in her career.

The conventional practice has been to tell freshmen not to be too concerned about their grades unless they can't pull their GPA up during their second semester, and to suggest they be more selective with their class schedules, balancing one or two difficult subjects with several easier courses each term, Ward Hood said.

"There are some things that suggest that we need to re-examine what we're doing and learn what's really going on with these students so we can personalize or target our interventions. Maybe what we're giving them is a handful of forks when what they really need is a spoon," Ward Hood said.

A first step toward developing effective interventions is for scholars to identify the underlying factors that may be negatively affecting these students' first-semester grades and ultimately their prospects of graduating, the researchers suggest.

"First-semester GPA is the proverbial canary in the coal mine," Gershenfeld said. "This research shows the need to intervene for students with a first-semester GPA below 2.33. At a time when great attention is focused on the graduation rates of underrepresented students, here is valuable evidence of how we can make a difference."


Social work professor Min Zhan, Gershenfeld's thesis adviser, also was a co-author on the study.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.