The effects of learning analytics hint system in supporting students problem-solving

Abstract

This mixed-method study examined the impacts of a learning-analytics (LA) hints system on middle school students’ problem-solving performance and self-efficacy (SE). Students in condition A received the LA hint system, students in condition B received a static hint system that contains the same set of hints but without the LA mechanism, condition C was a control group that no hints were provided. The statistical results showed that the problem-solving SE for students who engaged with the LA hint system improved significantly. Student interviews revealed that real-time supports and in-time positive feedback played key roles in supporting their SE growth. Moreover, student-generated quantitative and qualitative log data were collected for interpreting the research outcomes. The quantitative logs provided an in-depth examination of problem-solving strategies across the conditions while the qualitative logs provided another perspective to understand students’ problem-solving status. Implications for future implementation of LA-hint system in virtual PBL environments were provided.