Gender and AI in Flipped Classrooms: Unpacking the Impact of Chatbots on Engagement, Self-Efficacy, Writing Performance and Technology Affinity

The integration of artificial intelligence (AI) in flipped classrooms is reshaping foreign language writing instruction, yet its impact on student engagement, self-efficacy, writing performance, and technology affinity remains underexplored. This study investigates the effectiveness of AI chatbots in enhancing flipped learning by examining their influence on engagement, self-efficacy, writing performance, and Affinity for Technology Interaction (ATI), while controlling for gender. A quasi-experimental design involving 252 Indonesian English majors was employed, with participants divided into an experimental group (flipped learning + AI chatbot) and a control group (flipped learning only). Structural Equation Modeling (SEM) and Analysis of Covariance (ANCOVA) revealed that AI-enhanced flipped learning significantly improved self-efficacy, writing performance and engagement, though ATI declined over time. Gender differences emerged, with males exhibiting stronger cognitive engagement and self-efficacy, while females benefited more in emotional engagement and writing performance. These findings underscore the need for balanced AI-human interaction to optimize learning outcomes. This study contributes to emerging pedagogical models by highlighting the nuanced role of AI in language education and emphasizing inclusive, adaptive instructional design. Future research should explore diverse learner populations, long-term effects, and AI instructor proficiency to maximize the efficacy of AI-integrated flipped learning.