Trust Calibration in Clinical Decision Support: A Cognitive Framework for Uncertainty Representation

This paper models trust calibration in Clinical Decision Support Systems (CDSS) for mental health referrals through the development of the Cognitive Uncertainty Calibration Framework (CUCF). Existing research on CDSS primarily emphasizes system transparency and technical accuracy, often overlooking how clinicians cognitively process uncertainty. This study examines the mechanisms shaping how clinicians perceive uncertainty and calibrate trust in CDSS recommendations.CUCF models trust calibration through three interconnected dimensions: uncertainty perception and trust formation, adaptive trust regulation, and uncertainty-guided decision support. Unlike traditional models that treat trust as static, CUCF conceptualizes trust calibration as a dynamic process influenced by dual-process reasoning, cognitive biases, and structured uncertainty representation. We derive design considerations for uncertainty-aware CDSS, including cognitive-driven uncertainty visualization, bias mitigation strategies, and adaptive trust mechanisms. This study advances cognitive IS by integrating psychological principles into CDSS design, providing practical guidance for uncertainty-aware decision support and establishing a foundation for future empirical validation.