The Relevance, Reliability, and Validity of the National Graduate Admission Test (N- GAT) as an Admission Tool into Graduate Programs in Ethiopia

Abstract

Addis Ababa University introduced the Graduate Admission Test (GAT) for the first time in 2020 as ascreening aptitude test for candidates interested in applying for their postgraduate studies. Later, thiswas scaled up as a national GAT by the Ministry of Education in October 2023 and became amandatory screening test for all applicants to graduate programs in all public and privateuniversities. The need for the GAT arose due to concerns over the quality of graduate programsoffered by various universities in Ethiopia. The objectives of the study were 1) to explore the reliabilityand validity of the GAT; 2) to analyze the factors associated with the test, and 3) to suggest futureactions that could improve the quality of the test. Data were collected from the archives of the entiretest takers who sat for the national GAT in October 2023. The entire cohort of test takers, totaling34,520 candidates, had filled in their background information and their undergraduate GPAs, whichwere linked to their scores on the national GAT. Descriptive statistics, non-parametric tests (such aschi-square tests), and inferential statistics (such as t-test and analysis of variance) were used toanswer the research questions. Findings of this study showed that most candidates (about 82.4%)came from four regions; that is, from Oromia (28.5%), Addis Ababa (26.2%), Amhara (17.5%), andSNNPR (10.2%), while only about 17.6% came from the other regional states. Besides, there weresignificantly higher numbers of male (74.7%) than female (25.3%) test takers in all locations who satfor the GAT administration in October 2023. Scores on GAT were positively associated withcandidates’ socioeconomic background, where those who had internet access in their homes, forinstance, and those who had better undergraduate CGPA scored higher on the GAT. Furthermore, theinteraction of gender and undergraduate CPA was also explored using linear regression analysis, andit was found that female candidates who had higher undergraduate CGPA were more likely to scorehigher on GAT as compared to their male counterparts. Furthermore, as a screening tool, the GATwas found to be highly reliable, with moderate predictive validity for candidates who joined the MScand MA programs, but not for those who joined the PhD programs. Results were discussed andrecommendations forwarded.