Trust Calibration Between Human Experts and Autonomous Decision Technologies

Lucas Ferreira

Trust Calibration Between Human Experts and Autonomous Decision Technologies

Author(s)	Dr. Lucas Ferreira
Country	United States
Abstract	Autonomous decision technologies increasingly support experts in environments where decisions must be made rapidly despite uncertainty, complexity, and incomplete information. Their effectiveness, however, depends not simply on system accuracy but on whether human experts rely on automated recommendations appropriately. Excessive trust can produce automation bias, complacency, and acceptance of incorrect recommendations, whereas insufficient trust can cause experts to reject reliable assistance and lose potential gains in accuracy and efficiency. This study proposes an Adaptive Trust Calibration Framework for Human–Autonomy Decision Support (ATCF-HADS) that aims to align expert reliance with the demonstrated capability of an autonomous system rather than maximize subjective trust. The framework integrates system-confidence information, recent reliability history, decision criticality, human–system disagreement, explanation support, expert override, and continuous performance feedback. A design-science methodology was combined with structured evidence synthesis and a reproducible scenario-based simulation. The simulation modeled 2,400 expert–technology decision episodes distributed equally across high-reliability, moderate-reliability, and degraded-reliability conditions. An uncalibrated interaction strategy characterized by relatively stable dependence on automated recommendations was compared with an adaptive calibration strategy in which reliance thresholds changed according to confidence, reliability, disagreement, and task criticality. Appropriate reliance increased from 63.9% under the uncalibrated interaction to 93.4% under the calibrated configuration. Overreliance when the autonomous recommendation was incorrect declined from 76.1% to 4.6%, while overall human–technology team accuracy increased from 77.9% to 93.3%. The largest performance advantage emerged when autonomous reliability deteriorated, demonstrating the importance of helping experts recognize changing system capability rather than encouraging constant levels of trust. These numerical findings are simulation outputs rather than observations from real professionals. The study concludes that effective human–autonomy collaboration requires dynamic, evidence-sensitive trust calibration supported by transparent uncertainty communication, expert competence, meaningful override authority, and continuous monitoring of both technological reliability and human reliance behavior.
Keywords	trust calibration; human–autonomy interaction; autonomous decision systems; human expertise; appropriate reliance; automation bias; decision support; uncertainty communication; human oversight; trustworthy artificial intelligence
Field	Engineering
Published In	Volume 1, Issue 1, January-February 2020
Published On	2020-02-13

View / Download PDF File

About AJAMIR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+1-872-302-3851	editor@ajamir.com

American Journal of Advanced Multidisciplinary Innovation and Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Trust Calibration Between Human Experts and Autonomous Decision Technologies

Share this