Digital Literacy, Test Anxiety, and Achievement in STEM Examinations: A Comparative Study of Computer-Based Testing and Traditional Paper-Based Assessment

Abstract

Assessment plays a pivotal role in evaluating learning outcomes in science, technology, engineering, and mathematics (STEM) education, where conceptual understanding and problem-solving skills are critical. With the growing adoption of digital technologies, computer-based testing (CBT) has emerged alongside traditional written (paper-based) examinations, prompting debates about their comparative effectiveness. This study investigates the impact of both formats on academic performance, cognitive engagement, test anxiety, and student perceptions in Nigerian tertiary institutions. Adopting a mixed-method design, quantitative performance and perception data were complemented by qualitative insights from focus groups. Findings revealed that written examination participants achieved marginally higher scores (M = 71.3) than their CBT counterparts (M = 68.5), with the difference reaching statistical significance (p = 0.006). While CBT offered efficiency and rapid feedback, it was associated with higher anxiety levels and limitations in assessing complex reasoning. Written examinations, by contrast, better supported extended responses, graphical representation, and multi-step problem solving. The study recommends a balanced, context-sensitive assessment strategy that integrates both formats to harness their respective strengths, improve equity, and optimize learning outcomes in STEM disciplines.