Introducing Synthetic Psychology: Validation Across Personality and Established Relationships.

Abstract

This research introduces Synthetic Psychology (SP), a novel research approach in which large language models are configured to generate responses as synthetic individuals with predetermined psychological profiles. Across four studies with 20 replications, we configured Llama 3.3 as 32 distinct Big Five personality profiles (N = 30,400 synthetic respondents). Studies examined factor structure, personality-affect-wellbeing relationships, and personality-cognition relationships across extensive measures including risk-taking, need for cognition, thinking styles, and cognitive failures. Synthetic data replicated established patterns with exceptional fidelity and unprecedented reproducibility. Meta-analytic correlations demonstrated appropriate Big Five structure and theoretically expected relationships (e.g., Extraversion-Positive Affect r = .661-.664, Openness-Need for Cognition r = .984). While some correlations exceeded typical human magnitudes, results demonstrate SP's potential for preliminary hypothesis testing and theoretical exploration, with the caveat that synthetic data should be seen as a complement to rather than a replacement for human validation in psychological research.