Logged In and Mindset On: Log Data Based Insights on Engagement and Learnings in Programming CBLEs

Abstract

Mindset has been shown to influence learners' motivation, persistence, and academic outcomes. While extensive research supports the benefits of a growth mindset in traditional classroom settings, little is known about how these beliefs shape learning in autonomous digital environments. This study explores the relationship between students' mindsets (growth vs. fixed) and their performance and engagement in a computer-based learning environment (CBLE) designed for Python programming, called PyGuru. Sixty-seven first-year undergraduates participated in a two-day workshop involving mindset surveys, pre- and post-tests, and three hours of CBLE interaction. Learners were grouped into fixed and growth mindset categories using a tertile split. Learning gains were computed via Hake’s formula; behavioural data from log traces were analysed using Mann-Whitney U tests with Benjamini-Hochberg correction. Growth mindset students demonstrated significantly higher learning gains despite similar baseline scores. They also engaged more selectively—performing fewer quizzes, readings, and assessments—whereas fixed mindset learners showed more frequent error-checking. Time-on-task did not differ significantly, indicating that engagement quality, not duration, distinguished the groups. These results highlight that mindset influences not only outcomes but also learning behaviours in CBLEs, even without teacher support. Findings support the integration of mindset-sensitive features into digital learning tools and call for future research across larger, more diverse samples to inform adaptive, growth-oriented learning design.