Exploring Gender-Associated Patterns of Politeness and Directness in EFL Students' Prompts to Generative AI
This study investigates how EFL students employ politeness and directness strategies in their prompts to Generative AI and explores the underlying motivations shaping these linguistic choices. While prior studies have examined gendered communication and prompt effectiveness separately, limited research has integrated these perspectives within AI-mediated interaction. Grounded in Politeness Theory, the CCSARP framework, and Genderlect Theory, this study adopts a qualitative design combining directed content analysis and thematic analysis. The dataset consists of twenty-six prompts generated by four Indonesian EFL students, supported by semi-structured interviews. Findings reveal that bald-on-record and conventionally indirect strategies dominate, reflecting a balance between efficiency and minimal politeness. Thematic analysis identifies four key factors influencing prompting behavior: politeness strategies, directness strategies, variation strategies, and perceptions of AI. Politeness is shaped by sociocultural norms and expectations of improved output quality, whereas directness is driven by efficiency and reduced cognitive effort. Participants also demonstrate adaptive prompting behavior, reflecting emerging AI literacy through iterative refinement. Importantly, no strong gender-based differences were observed, suggesting that individual preferences and task demands play a more significant role. The study highlights that politeness and directness function as complementary strategies in AI-mediated communication and offers implications for developing AI literacy in language learning contexts.