York University research led by Associate Professor Khaled Barkaoui shows that different English-language proficiency tests predict academic performance in various ways and offers guidance on supporting multilingual students.
The study, published in Language Testing, examined what insights two of the most commonly used English-language proficiency assessments - IELTS Academic and TOEFL iBT - offer in predicting how undergraduates perform in university studies.
Because universities around the world accept multiple tests as equivalent demonstrations of readiness, Barkaoui, who teaches in the Faculty of Education, wanted to determine whether IELTS and TOEFL function similarly in correlations with academic achievements. He did so by exploring new ground.
"There is a lot of research looking at the relationship between English-language proficiency (ELP) and academic achievement for international students in English-speaking universities, but most of it looks at this relationship in the first year or semester of study," he explains.
Barkaoui sought longer-term understanding of the correlation as it can influence the resourcees they need and receive. "Understanding these long-term effects has important implications for university admissions policies and policies regarding English-language support for international students," he says.
Barkaoui used archival data of 6,481 undergraduates over as many as 10 semesters, analyzing grade point averages alongside language assessments scores, prior scholarly performance and academic programs. He then examined how performance changed as their studies progressed in relation to the ELP test used for admission.
The results show that initial assessment scores do not always indicate learning outcomes in the same way. IELTS scores were more closely linked to first-semester grades and to how students' scholarly performance changed over time, with lower scores often corresponding to early dips and higher scores to more stable results. In contrast, TOEFL results were only weakly associated with academic results and did not reliably predict either first-semester grades or longer-term GPA trends, suggesting the two tests capture different aspects of language readiness.
This underscore that success is multi-faceted and shaped by more than language alone, says Barkaoui. In programs with heavier reading and writing demands, such as social sciences and humanities, proficiency scores were more strongly tied to grades. In quantitatively oriented fields, such as business and economics, the relationship was weaker or sometimes negative, pointing to the role of discipline-specific skills. Other factors - such as prior education experiences, program requirements and support networks - also contribute to how students navigate university.
For that reason, Barkaoui's conclusion points toward helping institutions continue to refine how and when academic and language-learning supports are offered. The study suggests that undergraduate who enter with lower English-language proficiency may benefit from services that extend across multiple semesters, and that those resources may need to differ over time.
The research highlights why understanding IELTS- or TOEFL-specific patterns matters. If universities understand how each test relates to learning experiences, staff supporting students can proactively tailor resources that foster success for all learners.
"Understanding the long-term effects of ELP on academic success in different fields of study can inform policies regarding the use of ELP tests for admission and English-language support for these students and when it needs to be provided - whether early in their studies or throughout," Barkaoui says.
This story was originally featured in YFile, York University's community newsletter.
