Leveraging Artificial Intelligence for Adaptive Algebra Instruction: Effects on Students' Self-Efficacy and Problem-Solving Skills in Northern Nigeria

This quasi-experimental study investigated the impact of Squirrel AI, an Artificial Intelligence (AI)-powered adaptive learning platform, on senior secondary school students’ self-efficacy and problem-solving performance in algebra in Northern Nigeria. The study employed a pretest and post-test control-group design using intact classes. A total of 180 Senior Secondary II students from four public co-educational secondary schools in Zaria Education Zone of Kaduna State participated, with 90 
students in each of the experimental and control groups. The experimental group received instruction through the Squirrel AI 
platform, which adapts learning pathways by diagnosing students’ misconceptions, adjusting content difficulty, and providing real-time feedback, while the control group was taught using conventional lecture methods. Two instruments were used for data collection: the Algebra Self-Efficacy Scale (ASES) and the Algebra Problem-Solving Test (APST). Data were analyzed using descriptive statistics and Analysis of Covariance (ANCOVA). Results revealed that students in the experimental group recorded significantly higher self-efficacy and problem-solving scores than those in the control group, with large effect sizes (partial η² = 0.180 and 0.298, respectively). These findings underscore the transformative potential of AI-powered instruction in enhancing mathematics outcomes in resource-limited contexts such as Northern Nigeria, where infrastructural challenges and educational 
disparities often hinder student achievement. The study concludes by recommending the integration of AI tools into the Nigerian secondary school curriculum and teacher training programs to foster personalized, data-driven, and equitable learning opportunities in mathematics.