Predicting nursing baccalaureate program graduates using machine learning models: A quantitative research study.

Area of Science:

• Nursing Education
• Educational Data Mining
• Machine Learning in Higher Education

Background:

• Nursing schools face challenges with high student attrition and low completion rates.
• Early identification of at-risk students is crucial for improving graduation outcomes.
• Machine learning offers advanced predictive capabilities compared to traditional statistical methods.

Purpose of the Study:

• To develop predictive models for undergraduate nursing student graduation within six years.
• To identify key factors associated with successful graduation using machine learning algorithms.
• To enable early intervention for students at risk of dropping out.

Main Methods:

• Utilized eight machine learning algorithms and a stacked ensemble method for model construction.
• Evaluated prediction accuracy at five distinct time points throughout the nursing program.
• Developed fourteen scenarios by combining different variable sets and time points.

Main Results:

• Graduation outcomes predicted with over 80% accuracy after one year of academic performance.
• Prediction accuracy reached 90% after the second year and 99% after the third year.
• Cumulative Grade Point Average (GPA) and nursing course GPA were the most significant predictors of graduation.

Conclusions:

• The study presents a data-driven system for tracking nursing student progress.
• This system can automatically assess dropout risk and inform targeted interventions.
• It enhances the capacity of nursing schools to implement strategic support services for student success.