Thesis Format

Integrated Article

The Predictive Validity of IELTS Scores: A Meta-Analysis

Tomlin Gagen, The University of Western OntarioFollow

Degree

Master of Arts

Program

Education

Supervisor

Faez, Farahnaz F.

Abstract

Thousands of education institutions worldwide rely on IELTS scores as criteria for accepting international students whose first language is not English. Individual studies have found varying degrees of strength of correlations and conflicting results between IELTS scores and academic success. These conflicting results were examined through a meta-analysis, while also investigating multiple moderating variables: research funding bias, individual skill scores, level, field, & country of study, presence of additional English courses, and GPA timepoint.

Results from 18 studies show an approaching-small effect size of r =.227 for the relationship between IELTS scores and post-secondary GPA. The majority of macro skills (listening, writing, and speaking) do not reach the small effect size, however reading approaches it with an effect size of r =.215. Most moderator analyses were inconclusive owing to the small amount of studies, but potential differences from individual studies are examined and discussed.

Summary for Lay Audience

Increases in international student enrolment has led to an increased need to examine the language tests used to ensure they have the necessary English skills to succeed in university. One of the most popular of those tests is the International English Language Testing System, or IELTS. The present study is a meta-analysis, or synthesis of large amounts of already published research, into how IELTS scores used for admission relate to actual student scores once in post-secondary education. A total of 18 different studies are aggregated and examined to look for any evidence of bias in publishing from IELTS, as well as any stronger links for specific English skills (listening, reading, writing, and speaking). Additional examination is made into if the relationships are stronger for: different majors, different levels of study (graduate or undergraduate), different countries, presence of extra English courses, and results at different times.

Results show that the IELTS test has a relatively small predictive effect for GPA in post-secondary programs. No evidence of bias was found in publishing source, and overall scores were much more predictive than any one individual skill such as listening or writing, although reading shows a slight relationship. Due to limited sample sizes, few conclusions can be drawn from the presence of any extraneous variables.

Recommended Citation

Gagen, Tomlin, "The Predictive Validity of IELTS Scores: A Meta-Analysis" (2019). Electronic Thesis and Dissertation Repository. 6406.
https://ir.lib.uwo.ca/etd/6406