Tao Bi, Shakirova Araily, Anqi Lyu, Si Xiu
Students demonstrated a moderate level of digital literacy, with information usage and organization showing the highest scores, while metacognition and problem-solving were relatively lower. Engagement in AI-enhanced learning activities was positively and significantly related to all dimensions of digital literacy. SEM results further showed that AI-enhanced learning activities made a significant positive contribution to overall digital literacy.
INTRODUCTION: This mixed-methods field study examined generative artificial intelligence (GenAI) pedagogical readiness among university faculty in five participating universities in China and Kazakhstan, with particular attention to cross-sample differences and the effects of faculty development training.
METHODS: Baseline survey data were collected from 568 faculty members, including 340 from China and 228 from Kazakhstan. Readiness was assessed across seven dimensions: basic GenAI understanding, prompt design, AI-supported instructional design, AI-supported assessment, critical judgement, educational ethics, and disciplinary transfer. A quasi-experimental subsample of 160 faculty members compared structured GenAI prompt-task training with conventional GenAI-focused faculty development. Prompt logs were additionally coded to identify post-training faculty development needs.
RESULTS: At baseline, the Kazakhstani sample reported higher readiness across all seven dimensions, with the largest differences in disciplinary transfer, prompt design, and basic understanding. Measurement-invariance analysis supported cross-sample comparison of observed mean patterns. In hierarchical regression, the Kazakhstan coefficient decreased from B = 0.205 to B = 0.041 after prior GenAI use, recent AI-related training, multilingual resource access, institutional support, perceived permission, policy clarity, risk sensitivity, and disciplinary background were included. Post-test comparisons showed greater readiness in the structured prompt-task training group than in the conventional group, particularly for prompt design and critical judgement. Prompt-design gains remained above baseline at delayed testing. Prompt evidence linked these gains to objective specification, constraint setting, assessment criteria, faculty verification, and risk-control prompts.
DISCUSSION: The findings indicate that GenAI pedagogical readiness is shaped by faculty course tasks, institutional support, resource access, and training design. Structured prompt-task training may therefore provide a useful approach to strengthening course-level faculty readiness for pedagogical uses of GenAI.