2026/07/23 by Kit Ling LAU, Kit‐ling Lau, Yu Ka Wong +1
Computer Science · Medicine · Social Sciences · #AI in Service Interactions #Artificial Intelligence in Healthcare and Education #Writing and Handwriting Education
paper · doi:10.1111/ijal.70320
openalex publication_date 2026/07/23 · openalex created_date 2026/07/24 · openalex updated_date 2026/07/29
ABSTRACT While the benefits of generative artificial intelligence (GenAI) in supporting various cognitive processes in writing have been documented, most research has examined single stages of the writing process and focused on university students. This study addresses this gap by developing a domain‐specific GenAI chatbot and embedding it within a process‐based writing instruction (PBWI) program to explore its effectiveness and challenges in supporting secondary students’ Chinese writing across the three cognitive stages. A total of 100 first‐year secondary students participated in three conditions: the experimental group received PBWI with GenAI chatbot support, the control group received PBWI without GenAI, and the baseline group continued product‐based instruction. Data from writing tasks, questionnaires, and interviews were analyzed using a mixed‐methods approach. Results indicated that students reported benefits of the GenAI chatbot in supporting their writing across the three cognitive stages, as well as in both lower‐ and higher‐level aspects of writing. However, their performance outcomes did not significantly exceed those of the control group, likely owing to the added cognitive load of managing both PBWI and GenAI use. Students nonetheless expressed a strong intention to continue using the chatbot and reported enhanced confidence and motivation. These findings highlight both the promise and limitations of integrating GenAI into writing instruction and provide insights into the design of pedagogical frameworks that combine PBWI with tailored GenAI tools at the secondary level.