Mandel-AI Effect:Proposed measurement of the Mandela Effect induced by artificial intelligence in digital social networks.

Abstract

This study proposes and validates a pioneering psychometric instrument to measure the “Mandel-AI Effect,” a contemporary manifestation of the Mandela Effect enhanced by artificial intelligence (AI) in digital social networks. Following a rigorous theoretical design and expert judgment validation, the instrument was developed and tested through exploratory factor analysis using principal component extraction with oblique rotation (oblimin), complemented by confirmatory factor analysis. The final scale comprises three dimensions: AI Presence, Reality Distortion, and the Mandel-AI Effect itself. The study was conducted with a sample of 243 centennial-generation university students from southern Tamaulipas, Mexico, aged between 18 and 28 years, all active users of digital social networks. Statistical analyses confirmed the factorial structure of the model, with high levels of internal consistency, convergent validity, and discriminant validity. Cronbach’s alpha coefficients for each dimension were excellent, indicating strong reliability. The findings show that AI not only mediates content selection and viralisation processes but also structurally influences the configuration of cognitive and cultural frameworks underpinning the social construction of collective memory. This influence can facilitate the implantation of false memories, induce epistemic confusion, and blur the boundaries between authentic and artificial content. The validated instrument offers a robust tool for future research on the cognitive and social effects of AI, contributing to the emerging field of digital psychology and providing a methodological basis for the systematic study of one of the most pressing epistemic challenges of the digital age.