PERCEIVED EFFECTS OF AI TUTORS ON PROGRAMMING SKILL DEVELOPMENT: A CROSS-SECTIONAL SURVEY OF BSCS STUDENTS

Abstract

Artificial intelligence (AI) tutors have become increasingly common in university programming courses, yet persistent debate remains regarding whether frequent use translates into authentic skill development. While these tools offer immediate debugging and syntax support, there is a concern that they may serve as shortcuts rather than learning enhancers. Using a quantitative, cross-sectional survey design, data were analyzed from participants using a five-point Likert instrument containing 30 items. The questionnaire demonstrated strong internal consistency, capturing demographics, usage intensity, and multiple perception dimensions. Descriptive results indicated moderate-to-high AI usage and moderate-to-high perceived skill development. However, Pearson correlation showed a statistically significant relationship between usage frequency and perceived skill gains, and ANOVA results confirmed significant between-group separation. The evidence suggests that student perception is multidimensional and not explained by frequency alone. Findings reinforce the need for guided AI integration and process-focused pedagogy to ensure AI functions as a learning amplifier. This study provides a context-specific data baseline for BSCS programs in Pakistan, offering practical insights for curriculum redesign and future longitudinal research aimed at optimizing AI-human collaboration in computer science education.