Apriori Algorithm Reveals Learned Helplessness Patterns in Math Tutoring
A recent study featured on arXiv (2604.26237) utilized the Apriori algorithm to investigate behavioral interaction trends related to learned helplessness (LH) in logs from a mathematics tutoring system. The researchers analyzed interaction data across three key areas: LH levels (low vs. high), the presence of system-based interventions (with vs. without), and problem-solving results (solved vs. unsolved). Findings indicated that the most common behavior linked to unsolved problems was skipping questions without hints, while persistence, such as not skipping, was less prevalent. Low-LH students showed a stronger connection between problem-solving and not skipping, alongside positive correlations between hint usage and solved problems. Conversely, high-LH students displayed more avoidance behaviors, with skipping closely associated with unsolved outcomes. Additionally, those without interventions frequently skipped without hints, implying that interventions could reduce avoidance behaviors. This study highlights the effectiveness of association rule mining in educational data analysis to pinpoint maladaptive learning behaviors.
Key facts
- Study applied Apriori algorithm to analyze learned helplessness in math tutoring logs.
- Data examined across LH level, intervention, and outcome dimensions.
- Skipping without hints was the most frequent pattern for unsolved outcomes.
- Low-LH students showed stronger links between solving and not skipping.
- High-LH students showed avoidance patterns with skipping tied to unsolved outcomes.
- Students without intervention had highest frequency of skipping without hints.
- Study published on arXiv with ID 2604.26237.
- Research demonstrates utility of association rule mining in educational data mining.
Entities
Institutions
- arXiv