AI-Reality Discrimination Assessment (ARDA-20): A Standardized Psychometric Instrument for Evaluating Individual Capacity to Discriminate AI-Generated Content from Authentic Reality

The AI-Reality Discrimination Assessment (ARDA-20) is a standardized psychometric instrument designed to evaluate an individual’s capacity to distinguish AI-generated content from authentic reality across five theoretically distinct cognitive domains: Factual Accuracy Detection (FAD), Perceptual Boundary Awareness (PBA), Epistemic Calibration (EC), Emotional & Relational Reality Testing (ERT), and Meta-Cognitive Awareness (MCA). The instrument comprises 20 dichotomous items and yields a composite AI-Reality Dissociation Index (ARDI) on a continuous scale from 0.00 to 1.00. Grounded in Reality Monitoring Theory (Johnson & Raye, 1981), Epistemic Vigilance (Sperber et al., 2010), and Epistemic Calibration research (Pennycook & Rand, 2019), the ARDA-20 provides criterion-referenced interpretation thresholds and domain-specific sub-indices. The instrument is intended for educational, research, and media literacy applications and does not constitute a clinical diagnostic tool.