Mechanisms linking epistemic curiosity and learning performance: The multifaceted role of mind wandering from trait and state analysis

Epistemic curiosity (EC) positively predicts mind wandering (MW) and learning performance; however, MW negatively impacts learning performance. Clarifying MW's role in EC's impact on learning performance is essential. This study examined how distinct MW dimensions shaped the EC-learning performance relationship at trait and state levels, and explored whether autonomous choice enhanced EC and reduced MW occurrence. Participants in Study 1 ( N = 318) and Study 2 ( N = 80) were Chinese college students. Study 1 used questionnaires to investigate how distinct dimensions of MW influence the EC-learning performance relationship at the trait-level. Based on Study 1, Study 2 examined these relationships at the state-level through video lecture learning and evaluated the effectiveness of enhancing students’ EC by autonomous choice. Study 1 revealed that EC was positively associated with intentional MW (iMW) and learning performance, whereas both iMW and unintentional MW (uMW) negatively affected learning performance. Significantly, iMW mediated the relationship between EC and learning performance, with a masking effect observed. Study 2 yielded partially different findings compared to Study 1. During science learning (state level), EC negatively correlated with the intentionality, frequency, and depth dimensions of MW. However, only MW depth mediated the relationship between EC and learning performance in science learning. EC positively predicts learning performance at both state and trait levels. However, the relationship between EC and MW varies depending on context and MW dimensions. Furthermore, the distinct dimensions of MW play different roles in the relationship between EC and learning performance. • EC enhances learning performance at both the trait and state levels. • EC's impact on MW varies with the trait-state distinction and MW dimensions. • EC promotes learning performance by reducing the depth of the MW. • Compared with other dimensions, MW depth has a more substantial impact on learning. • Fostering EC through autonomous choices could optimize learning performance.