Predictive analytics in gamified education: A hybrid model for identifying at-risk students

Area of Science:

• Educational Technology
• Machine Learning in Education
• Student Support Systems

Background:

• Student disengagement and dropout pose significant challenges in educational settings.
• Gamified learning environments generate rich behavioral data for student analysis.
• Traditional assessment methods may not adequately identify at-risk students early.

Purpose of the Study:

• To develop and validate a hybrid predictive model for accurately identifying at-risk students in gamified education.
• To leverage machine learning ensembles for enhanced student risk assessment.
• To provide educators with an effective tool for timely academic interventions.

Main Methods:

• Integration of logistic regression, decision trees, and random forests into an ensemble model.
• Analysis of student data including academic performance, participation, and task completion from gamified platforms.
• Development of a machine learning-powered student monitoring system.

Main Results:

• The hybrid ensemble model demonstrated superior performance in identifying at-risk students compared to individual classifiers.
• The model accurately predicts students requiring intervention based on gamified learning data.
• Early detection of at-risk students was achieved, facilitating timely support.

Conclusions:

• Hybrid machine learning models offer a robust approach to at-risk student identification in gamified learning.
• The developed system provides actionable insights for educators to implement targeted interventions.
• This approach enhances the efficiency and effectiveness of student support services.