Impact of AI on Students' Performance in Economics in Kaduna Metropolis Secondary Schools

Abstract

This study investigated the effect of Artificial Intelligence (AI) tools on the academic performance of Senior Secondary School students in Economics in selected schools within Kaduna Metropolis, Kaduna State, Nigeria. Adopting a descriptive correlational survey design, data were collected from five hundred (500) Senior Secondary School Economics students, drawn from a purposively selected set of public and private schools within the metropolis (identities withheld in line with the confidentiality undertakings given to participating institutions), using a structured questionnaire that measured demographic characteristics, access to and frequency of AI tool use, perceptions of the positive effects and challenges of AI tools, and academic performance in Economics expressed as a percentage score. Data were analysed using descriptive statistics, Pearson product-moment correlation, independent-samples t-tests, one-way analysis of variance (ANOVA), and multiple linear regression. The results showed that 83.6% of respondents had access to AI tools, with ChatGPT being the dominant tool used (50.6%). Students held moderately positive perceptions of the benefits of AI tools (overall mean = 3.01 out of 5.00) and moderate levels of concern about associated challenges (overall mean = 3.01 out of 5.00). A strong, positive, and statistically significant correlation was found between students' perceived positive effect of AI tools and their academic performance in Economics (r = 0.590, p < .001), while the correlation between perceived challenges and performance, though significant, was weak (r = 0.121, p = .007). Academic performance differed significantly by sex, favouring females (p = .026), and rose monotonically across categories of AI use frequency, from Never (M = 41.23%) to Very Often (M = 69.58%), F(4, 495) = 97.52, p < .001, but did not differ significantly between public and private school students (p = .118). A regression model combining both perception variables explained 35.0% of the variance in academic performance, with perceived positive effect as the dominant predictor. The study concludes that AI tools exert a substantively positive effect on students' academic performance in Economics, particularly with frequent and purposeful use, although moderate levels of concern about over-dependence and information accuracy indicate that these risks require deliberate pedagogical attention. It is recommended that Economics teachers structurally integrate AI tools into instruction, that schools develop clear guidelines for their use, and that AI literacy and information-verification skills be explicitly taught alongside subject content.