This study examines how student characteristics, item properties, and their interactions influence mathematics literacy using data from 7,250 Turkish students in the Programme for International Student Assessment (PISA) 2022. Employing Explanatory Item Response Modeling (EIRM), the analysis incorporated individual-level factors such as response styles (acquiescent, midpoint, extreme, and disacquiescent) and gender, along with item-level variables including format, cognitive demand, and response time behavior (solution-oriented vs. rapid guessing). Results indicate that acquiescent and midpoint response styles are positively associated with performance, while extreme response style is negatively linked to performance. Male students outperformed females. Open-ended and reasoning-level questions were more challenging than multiple-choice or other cognitive-demand items. Significant interactions emerged between response styles and time-related behaviors. Beyond these empirical findings, this study contributes conceptually by positioning response styles, traditionally associated with attitudinal measures, within cognitive performance. The results carry implications for the design of large-scale assessments, highlighting the importance of accounting for both cognitive and behavioral factors in interpreting test performance.
Dibek et al. (Sun,) studied this question.