Answers are grounded in this paper and cite their sources.
This study examines prompt literacy among future computer science teachers in the context of solving and debugging programming tasks with ChatGPT. It focuses on students' perceptions of prompt literacy, their understanding of responsible AI use, documented practices of verifying and modifying AI-assisted solutions, and patterns of academic performance. A descriptive multi-method study was conducted during a 16-week programming course involving 113 future computer science teachers at Kh. Dosmukhamedov Atyrau University, Kazakhstan. Data were collected through questionnaires, student programming submissions, ChatGPT-use disclosure notes, reflective comments, code-similarity screening, and course-grade records. Quantitative data were analyzed descriptively, while qualitative content analysis was used to identify recurring patterns in students' reported use, verification, and revision of AI-generated solutions. The findings were compared across data sources to address the research questions without testing causal effects or participant-level associations. The 20-item Prompt-Literacy Perceptions Questionnaire demonstrated high overall internal consistency (Cronbach's α = 0.94). Students generally reported positive perceptions of ChatGPT for code generation, debugging, and explaining programming logic. They also recognized the importance of verifying AI-generated outputs and disclosing substantial AI assistance. However, students' documentation of verification and modification practices was frequently general rather than detailed. Pairwise code-similarity screening identified relatively few submission pairs exceeding the predefined 50% similarity threshold. Mean academic scores were numerically close across practical assignments and independent control tasks, although differences in assessment conditions and grading criteria limited direct comparisons. The findings highlight the importance of integrating prompt literacy, critical verification, and transparent documentation of AI use into computer science teacher education. Positive perceptions of responsible ChatGPT use did not necessarily correspond to detailed documentation of verification and revision practices. Code-similarity results cannot independently establish whether students modified AI-generated solutions. Furthermore, the descriptive single-cohort design does not support conclusions about causal effects or improvements in prompt-literacy competence. The study emphasizes the pedagogical value of making students' reasoning, verification, and responsible AI-use practices explicit during programming instruction.
No takes yet. Share an insight, caveat, or question.
RAKHMETOV et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: