Reading comprehension across paper and screen: performance, processes, and cognitive predictors in different tasks

Abstract

ABSTRACT Going beyond the study of differences between media, the present study investigated the extent to which individual (i.e. reading comprehension, updating, and verbal/nonverbal reasoning) and contextual (medium) factors contribute to explaining variability across performance (accuracy), process measures (time-on-task and lookbacks), and metacognitive calibration (calibration bias) obtained through different comprehension tasks (inferential, cloze, and error detection tasks). The sample of 197 students (grades 4–7) completed the paper-based and computer-based comprehension tasks on small narrative texts, differentiated by grade. Results from a model-comparison approach using general linear mixed models showed that while performance measures are better explained by considering individual differences, time-on-task and lookback measures are largely dependent on the medium. The results on calibration bias were consistent with previous evidence of overconfidence in digital comprehension. Nonetheless, results varied across tasks, with the cloze task showing no effect of medium. These findings suggest that the impact of reading medium is more evident on process and metacognitive measures, while individual differences should be taken into account when evaluating comprehension. This underscores the need to contextualize the use of different media according to the specific demands of the individual and the assessment setting.