Comparing AI chatbots and web-based synchronous collaborative writing tools, leveraging group cohesion, motivation,...

Area of Science:

• Educational Technology
• Applied Linguistics
• Human-Computer Interaction

Background:

• Growing adoption of Generative AI (GenAI) in language education.
• Gaps in psychological research on AI-supported collaborative writing, motivation, and self-efficacy in group settings.
• Lack of direct comparisons between AI chatbots and traditional tools in real-time, in-class collaborative writing.

Purpose of the Study:

• To examine the effects of GenAI chatbots versus traditional web-based resources (TWRs) on group cohesion, motivation, and writing performance.
• To investigate these effects during face-to-face synchronous collaborative writing (SCW) tasks among L2 learners.

Main Methods:

• Sixty-three L2 learners were randomly assigned to use either an AI chatbot (Kimi) or TWRs (dictionaries, search engines) for SCW tasks.
• Mixed-methods approach including screen recordings, writing assessments, motivation surveys, and cohesion notes.
• Analysis based on Group Cohesion Theory (GCT).

Main Results:

• The AI group demonstrated significantly faster writing and proofreading times and achieved higher essay scores.
• AI group reported substantially greater post-task motivation compared to the TWR group.
• Qualitative data indicated AI streamlined communication and reduced information overload but could limit discussion and encourage over-reliance.

Conclusions:

• GenAI chatbots can enhance L2 learners' writing performance and motivation in SCW tasks.
• AI appears to strengthen task cohesion by improving coordination and performance.
• AI's impact on social cohesion and deep peer interaction requires further investigation, as challenges like reduced discussion and over-reliance were noted.