Metacognitive Calibration of Understanding
Year: 2026
Type: preprint
Abstract: The ability to evaluate one's own understanding is fundamental to learning, professional reasoning, and effective decision-making. However, subjective confidence does not always correspond to demonstrated comprehension. Individuals may experience a strong sense of understanding while remaining unable to explain underlying mechanisms, apply knowledge to unfamiliar situations, or recognize errors. The increasing availability of generative artificial intelligence (AI) introduces additional challenges because coherent explanations and readily available assistance may influence perceived understanding without necessarily producing equivalent improvements in independent competence. This article examines metacognitive calibration of understanding as the correspondence between subjective judgments of comprehension and independently assessed cognitive performance. Building on research in metacognition, metacomprehension, cognitive offloading, and human–AI interaction, it proposes the Triangulated Understanding Calibration Framework (TUCF). The framework distinguishes perceived understanding, demonstrated unaided comprehension, assisted task performance, and delayed transfer. A multidimensional assessment approach is proposed, incorporating factual comprehension, explanatory understanding, transfer, error detection, and confidence calibration. The article also differentiates calibration bias, metacognitive discrimination, and the potential misattribution of AI-supported performance to independent human competence. Six testable hypotheses and a proposed experimental design are presented. The framework is conceptual and has not yet undergone empirical validation. Its intended contribution is to support more precise investigation of how individuals evaluate their own understanding when cognitive work is increasingly mediated by artificial intelligence.
Author: Shaban, Roman
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7228486690
