“Even Though I Went Through Everything, I Didn’t Feel Like I Learned a Lot”: Insights From Experiences of Non-Computer Science Students Learning to Code

Research Background and Issues

• Identified Problems or Challenges: The paper explores the frustrations and satisfaction points encountered by non-computer science (CS) students in programming courses. These students often face obstacles due to insufficient course design, resources, and support tools, with many courses failing to meet their cognitive goals and motivations.
• Significance: Programming has expanded into numerous non-computer science fields, such as engineering, biological sciences, and supply chain management, becoming a critical component of university education. Understanding the learning experiences of non-CS students can help optimize programming course design to meet the needs of a broader student population.
• Research Motivation and Related Work: Previous studies have primarily focused on the learning experiences of computer science students or those in specific fields, lacking comprehensive research across universities and disciplines. Additionally, existing courses are often designed based on the intuition of computer science scholars, neglecting evidence-driven practices. The study also references the concept of "conversational programmers" proposed by P. Chilana et al., suggesting that some students aim not to become professional programmers but to understand programming languages to enhance their work capabilities.

Solution

• Proposed Methods or Solutions: The authors analyzed semi-structured interviews with 12 non-CS students from different universities and disciplines, identifying satisfaction and frustration points in course design and extracting recommendations for improvement.
• Innovations:
  • Emphasizing "Cultural Resonance" and the "Power Principle" to stimulate students' interest and motivation in learning programming.
  • Exploring programming education from a multidisciplinary perspective, highlighting non-CS students' preference for hands-on learning and domain-specific applications.
  • Advocating for the use of Generative AI to assist non-CS students in learning programming.
• Implementation Steps and Techniques:
  1. Collecting student learning experience data through semi-structured interviews.
  2. Organizing and analyzing data using Reflexive Thematic Analysis.
  3. Identifying course and environmental factors contributing to student satisfaction and frustration.
  4. Proposing specific course design recommendations and tool improvement directions for non-CS programming education.

Research Outcomes

• Specific Findings:
  • Non-CS students are primarily frustrated by low cultural resonance in classroom design, unreasonable difficulty levels, and limited resource support.
  • Encouraging diverse learning pathways and domain-specific course content can enhance students' interest and engagement in programming education.
  • Students particularly value hands-on practice and project-driven learning methods, which can significantly improve their learning outcomes.
• Comparison with Existing Solutions and Advantages:
  • By emphasizing cross-disciplinary design, the study highlights the construction of a student-centered course framework for non-CS students, contrasting with traditional computer science courses.
  • The research not only provides theoretical foundations (e.g., constructivism and cultural resonance) but also offers specific practical recommendations.
• Experimental or Evaluation Results:
  • Although the study sample includes only 12 students, their interviews reflect representative insights into the frustrations and expectations of non-CS students regarding programming courses, such as their preference for active learning and project-based teaching models.
  • Students expressed positive attitudes toward specific tools like Jupyter Notebooks and AI-based tutors, indicating that improved resource tool design could help reduce learning difficulties.
• Limitations and Future Directions:
  1. The study sample is limited to R1 universities in the United States, which may not fully represent the experiences of students from other educational backgrounds.
  2. The data includes a low proportion of non-STEM students, necessitating further research into the programming needs of students from non-technical disciplines.
  3. The current study primarily involves interview analysis, lacking field validation of the effectiveness of actual course designs and tools.

Conclusion

The paper reveals the challenges and potential opportunities faced by non-CS students in programming education, proposing innovative course and tool design recommendations. This study not only enriches the perspective of programming education research but also provides an inspiring framework and foundation for evidence-based improvements in programming education for non-CS students in the future.