開かれたパンドラの箱：撤回された論文を通じた査読者の修辞的感受性の育成

Retractions issued for misconduct offer a unique window into how questionable research is rhetorically constructed and made to appear credible. This study investigates how engaging with retracted articles can serve as a pedagogical tool for reviewer training, with particular attention to the rhetorical mechanisms through which unreliability is performed. Twenty STEM doctoral researchers analyzed self-selected retracted papers using guided critical-reading questions to identify problematic rhetorical features. Across the analyses, five recurring issues emerged: intertextual falsification, methodological opacity, rhetorical inconsistency, rhetorical overstatement, and terminological distortion. The findings indicate that this approach has the potential to raise doctoral students' rhetorical sensitivity by enabling them to detect subtle markers of unreliability and to adopt a more evaluative rhetorical stance toward scholarly texts. Retracted articles thus can provide an authentic pedagogical resource for developing reviewer rhetorical sensitivity within doctoral education.