第二语言写作
第二语言
反射(计算机编程)
心理学
任务(项目管理)
计算机科学
感知
数学教育
情感(语言学)
突出
书写系统
合作写作
写作评估
语言学
写作过程
学术写作
第一语言
生成语法
维数(图论)
书面语
英语作为第二语言
第二语言习得
多媒体
任务分析
定性性质
教育学
应用语言学
定性研究
公立大学
作者
Lanxuan Xie,Xue Wang,Lianjiang George Jiang,Yi Zhao
摘要
There have been mixed perceptions about the effect of digital multimodal composing (DMC) on learners' second language (L2) writing performance. This debate is reshaped as generative AI (GenAI) transforms both text‐based writing and DMC processes, which may affect their impacts on L2 writing performance. This 14‐week quasi‐experimental study compares students' L2 writing performance development in GenAI‐assisted DMC, GenAI‐assisted text‐based writing, and text‐based writing conditions. A total of 161 Chinese university students from three intact classes were randomly assigned to three instructional groups: GenAI‐assisted DMC (54 students), GenAI‐assisted text‐based writing (57 students), and a text‐based writing group (50 students). The DMC group received instruction in creating two digital posters, whereas the other two groups were guided to write two 150‐word essays. Data were collected from pre‐ and post‐writing tests, semi‐structured interviews, screen recordings of the writing process, and students' written reflections. Results show students in the GenAI‐assisted DMC group significantly outperformed the other two groups in both overall writing scores and content dimension scores. Analysis of qualitative data indicates that the GenAI‐assisted DMC group engaged in more reflection and revision, which may be related to three salient features of the DMC task when compared to text‐based writing: (1) reducing pressure and promoting engagement in GenAI‐assisted L2 writing, (2) prompting critical engagement with GenAI output based on audience and genre awareness, and (3) encouraging students to learn from each other's L2 expressions. This study demonstrates how DMC can mitigate concerns about superficial GenAI use, promoting deeper reflection and engagement with AI‐generated content and enhancing L2 writing development.
科研通智能强力驱动
Strongly Powered by AbleSci AI