AI-INTEGRATED REFLECTIVE LEARNING DESIGN AND PRE-SERVICE TEACHERS’ COMPETENCY DEVELOPMENT: A MIXED-METHODS STUDY

Onnipa Thaithae Pansamut; Phatchara Phayakka; Xiaozhuan Wang

doi:10.32890/mjli2026.23.2.3

Authors

Onnipa Thaithae Pansamut Walailak University, School of Education & Research Center of Geosocial and Cultural Research for Sustainable Development, Walailak University, Nakhon Si Thammarat, Thailand, Thailand
Phatchara Phayakka School of Education, Walailak University, Nakhon Si Thammarat, Thailand, Thailand
Xiaozhuan Wang School of Humanities, Ningbo University of Finance and Economics, Zhejiang, China, China

DOI:

https://doi.org/10.32890/mjli2026.23.2.3

Keywords:

Generative artificial intelligence, pre-service teachers, learning design competency, reflective mediation, reflective scaffolding, AI as a co-thinking partner

Abstract

Purpose – This study examined the developmental change in pre-service teachers’ learning design competency and reflective thinking associated with their participation in an AI-integrated reflective learning design, and explored the reflective processes through which generative AI was used during instructional design. The study did not claim that the design would result in improvement, but it did provide a preliminary, within-group evidence and a process-oriented account.

Methodology – A convergent mixed-methods design was employed with 60 pre-service teachers enrolled in a learning design course. The structured learning design embedded generative AI within iterative cycles of design, reflection, and revision. Quantitative data were collected using a rubric-based learning design competency assessment and a 12-item reflective thinking questionnaire administered before and after participation, which was subsequently analysed using paired-samples t-tests. Qualitative data from 60 learning logs and 60 AI interaction logs (284 coded excerpts) were analysed using reflexive thematic analysis (Braun & Clarke, 2006, 2021). The two strands were merged through a joint display.

Findings – Scores on overall learning design competency and reflective thinking showed a significant increase from pretest to post-test, with the largest change in reflection-on-action. Qualitative findings indicated that participants were engaged with generative AI as a reflective co-thinking partner through cycles of evaluation, modification, and rejection of AI-generated suggestions. However, there were also admissions of confusion, over-reliance or resistance among a small number of respondents.

Significance – The study contributes to a proposed reflective-mediation account that conceptualises generative AI as a pedagogically mediated tool rather than an efficiency-driven technology. For practice, the findings inform curriculum designers in teacher education by showing how reflective scaffolds (structured learning and AI interaction logs) can be embedded to support professional judgment. Given the single-group design, the contribution is exploratory and practice-oriented and warrants confirmation through comparative research.