<![CDATA[The reliability of aptitude tests is critical for ensuring fair and valid selection of candidates for higher education. This study investigated the reliability of the Abstract Aptitude Sub-Scale of a developed and validated aptitude test for UTME selection in Nigeria. Specifically, the study examined both internal consistency and test–retest reliability of the sub-scale. The research adopted a descriptive survey design and involved a sample of 200 senior secondary school students drawn from public and private schools in Delta State using purposive sampling. Data were collected through the 25-item Abstract Aptitude Sub-Scale and analyzed using Kuder-Richardson Formula 20 (KR-20), Cronbach’s alpha, item-total correlations, and Pearson correlation coefficient for test–retest reliability. Findings revealed that the sub-scale demonstrated high internal consistency, with KR-20 = 0.87 and Cronbach’s alpha = 0.88. Item-total correlations ranged from 0.35 to 0.58, indicating that all items contributed positively to the measurement of abstract reasoning. Additionally, the test–retest reliability was found to be excellent, with Pearson’s r = 0.89 and a low standard error of measurement (SEM = 1.1), demonstrating temporal stability over a two-week interval. These results confirm that the Abstract Aptitude Sub-Scale is a psychometrically sound instrument, capable of consistently and accurately assessing abstract reasoning ability among UTME candidates. The study concludes that the sub-scale is suitable for high-stakes educational assessment and recommends its adoption for national selection purposes, with periodic review to maintain its reliability. Further research is suggested to examine the predictive validity of the sub-scale in relation to academic performance in tertiary institutions.]]>
