A Comparative Teaching Experiment of Restricted VS Unrestricted AIGC Programming Tools in Data Structure Course Design

DOI: https://doi.org/10.62517/jhet.202615505

Author(s)

Lingyan Yang*, Yong Wei, Meng Li

Affiliation(s)

School of Computer Science, Guangdong University of Science and Technology, Dongguan Guangdong, China *Corresponding Author

Abstract

Post-training questionnaires for a unified multi-file C++ data-structure project were distributed and processed for data cleaning, yielding 103 valid samples in total. The experimental group contained 54 students who were only permitted to use the VS Code CodeGeeX and Tabnine plugins, while the control group comprised 49 students granted full access to all generative-AI tools without usage restrictions. This study adopts a self-designed questionnaire consisting of five-point Likert scales, single/multiple-choice behavior items and open-ended text questions to carry out mixed quantitative and qualitative analysis. Reliability analysis, independent samples t-test and chi-square baseline balance test are conducted on scale data; meanwhile, frequency statistics of AI usage behaviors and thematic coding of open feedback are supplemented. Statistical results show that only the dimension of perceived project completion efficiency presents a significant inter-group difference, and the two cohorts hold distinct usage habits and learning pain points. This research provides empirical evidence for establishing hierarchical standardized AIGC management schemes in programming practical courses.

Keywords

AIGC; CodeGeeX; Tabnine; Data Structure Teaching; Comparative Experiment

References

[1]L. Evans and K. Miller, "Design Scaffolds for AI-Aided Introductory Programming," Computers & Education, vol. 268, Art. 104216, 2025. [2]L. Liu and S. Chen, "Controlled AIGC Application in Vocational Data Structure Courses," in Proc. 2024 IEEE International Conference on Computer Science and Education (ICCSE), 2024, pp. 456–461. [3] Y. Zhang, H. Li and J. Wang, "Research on the Application of Intelligent AI Coding Tools," IEEE Access, vol. 12, pp. 98765–98773, 2024. [4] Z. Huang and Y. Gao, "Balancing Learning Efficiency and Ability Cultivation," IEEE Trans. Educ., vol. 67, pp. 512–519, 2024. [5]M. Zhang and L. Yang, "Application and Practice of VS Code Plug-in AI Programming Tools," in Proc. 2025 IEEE International Conference on Computer Engineering Education (ICCEE), 2025, pp. 189–194. [6] H. Wang and M. Lin, "CodeGeeX-Assisted Programming Teaching Optimization," in Proc. 6th International Conference on Artificial Intelligence in Education (ICAIED), 2025, pp. 210–215. [7]X. Zhao, Y. Wu and Z. Zhang, "The Negative Impact of Unrestricted Generative AI," Comput. Educ. Res. (Chinese Journal), vol. 30, no. 2, pp. 89–96, 2023. [8]S. Wang, "Problems and Countermeasures of Unrestricted Generative AI," Mod. Educ. Technol. (Chinese CSSCI Journal), vol. 34, no. 5, pp. 78–85, 2023. [9]T. Chen and R. Xu, "Empirical Analysis of AI Dependency," in Proc. 2023 International Conference on Information System Educational Engineering (ISEE), 2023, pp. 321–326. [10]Q. Liu and H. Zhang, "Research on the Influence of Different AIGC Modes," J. Vocat. Educ. Technol., vol. 42, no. 4, pp. 66–73, 2024. [11]R. Chen, "Likert Scale-Based Empirical Research," in Proc. 5th International Conference on Statistical Analysis Educational Research, 2023, pp. 99–104. [12] W. Zhou, "Risk Prevention and Standardized Application Strategy," High. Educ. Forum, vol. 15, no. 6, pp. 33–38, 2023. [13]G. Peng and X. Cao, "Reforming Data Structures Education," in Proc. International Conference on Educational Information Technology, Atlantis Press, 2025, pp. 128–133. [14]P. Zhang and S. Liu, "Comparative Experimental Study of Restricted and Open AI," Int. J. Educ. Inf. Technol., vol. 9, pp. 41–48, 2024. [15]L. Wang, "Generative AI Integrated Teaching Reform," Computer Education (Chinese Professional Journal), vol. 11, pp. 62–66, 2022.