Adaptive vs. Non-Adaptive Test Mode Effects: Mediation of Intrinsic Cognitive Load and Effort

This study examined how computerized adaptive testing (CAT) and linear fixed testing (LFT) influence test-takers’ performance on an assessment of statistics knowledge, while considering the influence of factors such as test anxiety, cognitive load, and effort. Participants (N = 248; Mage=19.85 years) were randomly assigned to either test mode. Results showed that higher behavioral effort was associated with improved scores. Participants in the CAT condition scored lower on the test despite higher behavioral effort than participants in the LFT condition. They also reported greater intrinsic cognitive load, which was negatively associated with test scores. No significant differences were found in self-reported mental effort or test anxiety. Mediation analyses revealed behavioral effort and intrinsic cognitive load partially mediated effects of test mode on performance. Serial mediation analysis indicated cognitive load and behavioral effort are sequential mediators between test mode and test performance, with a small effect. Analyses conducted excluding less engaged participants reinforced the role of intrinsic cognitive load in predicting test outcomes. These findings highlight the need to optimize CAT designs to balance effort with cognitive demands, ensuring equitable assessment in achievement testing.