Personality Traits as Predictors of Computer Programming Performance Among Computer Science Students Using Recursive Feature Elimination

This study investigates the influence of personality traits on computer programming performance among computer science students. Programming proficiency is crucial for developing analytical and problem-solving skills in students, yet high failure rates in programming courses necessitate in early identification of struggling students. While previous research focused on programming behavior, this study uniquely investigates the role of personality traits in relation to programming performance. A survey questionnaire based on the HEXACO model was administered to computer science students, along with their programming course grades. The Recursive Feature Elimination (RFE) method was employed to identify relevant personality traits, revealing a notable relationship between these traits and programming performance. The optimum model, which included agreeableness, emotionality, extraversion, conscientiousness, openness to experience, and honesty-humility, attained 82% prediction accuracy. The findings suggest that combining specific personality traits indicates strong computer programming performance. This has implications for educational institutions, curriculum design, and industry recruitment. Integrating personality assessments into admissions, advising, and recruitment processes may aid in identifying and supporting students with programming aptitude. Further research is recommended to generalize findings across diverse contexts and explore the stability of these relationships over time.