Validity and Reliability Analysis of the Artificial Intelligence-Digital Life Balance Scale.

Abstract

This study aimed to develop and validate the Artificial Intelligence - Digital Life Balance Scale (AI-DLBS), a psychometric tool designed to assess the multidimensional impact of digital technologies and artificial intelligence (AI) on individuals' psychological, social, physical, and academic well-being. Utilizing ChatGPT-4, a novel AI-driven approach, the 40-item scale was constructed to measure five key dimensions: frequency and duration of digital device use, psychological and social effects, physical health impacts, academic performance, and technology access and dependency. Data were collected from three independent samples of university students in Turkey (N = 773, N = 325, N = 86) using convenience sampling. Exploratory and confirmatory factor analyses revealed a six-factor structure, explaining 60.83% of the variance, with acceptable model fit indices (e.g., RMSEA = 0.06, CFI = 0.90). The scale demonstrated strong internal consistency (Cronbach's α = 0.68-0.87) and test-retest reliability. The AI-DLBS offers significant potential for psychiatric research and clinical practice, enabling mental health professionals to evaluate technology-related risks, such as anxiety, social isolation, and dependency, and design targeted interventions, including digital detox programs. The innovative use of AI in scale development highlights both its efficiency and ethical challenges, such as data bias risks. Findings suggest the AI-DLBS is a reliable and valid tool for assessing digital life balance, with implications for global mental health research and policy-making. Future studies should validate the scale across diverse populations and cultural contexts.