Predicting student self-efficacy in Muslim societies using machine learning algorithms

Area of Science:

• Educational Psychology
• Machine Learning in Education
• Sociocultural Influences on Learning

Background:

• Self-efficacy is crucial for academic success and life outcomes.
• Research on self-efficacy predictors using machine learning is limited, especially in Muslim societies.
• This study addresses this gap by analyzing factors influencing student self-efficacy in these contexts.

Purpose of the Study:

• To identify key predictors of self-efficacy among secondary school students in Muslim societies.
• To leverage machine learning models for analyzing these predictors.
• To provide a data-driven foundation for educational interventions.

Main Methods:

• Empirical dataset of secondary school students in Muslim societies.
• Four machine learning algorithms (Decision Tree, Random Forest, XGBoost, Neural Network) were used.
• Predictors included demographic, socio-emotional, cognitive, and regulatory factors, with culturally relevant variables.

Main Results:

• Random Forest model demonstrated superior accuracy (R-squared and RMSE).
• Self-regulation, problem-solving, and sense of belonging were the most significant predictors.
• Gratitude, forgiveness, empathy, and meaning-making showed moderate influence; gender, emotion regulation, and collectivist-individualist orientation had minimal impact.

Conclusions:

• Machine learning effectively identifies self-efficacy predictors in diverse cultural contexts.
• Self-regulation and socio-emotional factors are critical for student performance and well-being.
• Findings support targeted educational interventions to enhance student outcomes in Muslim societies.