From Conceiving to Calibrating Requests: How Students Seek Feedback From Generative Artificial Intelligence in Language Interpreting

ABSTRACT While past studies have documented performance improvements resulting from student–generative artificial intelligence (GenAI) interactions, few have systematically examined the specific processes of feedback seeking or connected them to corresponding areas of improvement. To address this gap, the present study conceptualizes feedback seeking from GenAI as comprising two subprocesses: conceiving and calibrating requests. Twenty students learning language interpretation were invited to individually interact with GenAI for feedback. Two measures of feedback requests (feedback‐seeking forms and questionnaires) were linked to two outcome variables (interpretation quality gains and perceived feedback usefulness). The findings indicate that levels of feedback conception were more strongly associated with both quality gains and perceived usefulness than levels of calibration. Student interviews further revealed that deep planning and skillful calibration with GenAI stimulated reflection and revision, thereby consolidating their mastery of interpreting skills.