From latent constructs to networks: modeling high-dimensional social inferences in naturalistic settings

People make a variety of social inferences about others in daily life. Long-standing research suggests that these inferences are captured by a few latent dimensions (e.g., warmth and competence). However, prior work relied on constrained designs (e.g., face images, forced ratings), limiting ecological validity. Others suggest that social inferences are more complex than just a few dimensions could summarize but lack empirical support. Here, we conducted two pre-registered studies to test the high-dimensional properties of social inferences. To maximize generalizability, we computationally sampled diverse naturalistic videos and recruited U.S. representative participants (N = 1,598). Participants freely described people in videos using their own words. Cross-validation identified 25 latent dimensions which explained only 15% of the variance in the data. Alternatively, a sparse network model representing the unique correlations between inferences unreducible to shared latent constructs better represented the data. The network models informed the dynamics of naturalistic inferences, revealing what inferences people tended to make for the same individual and how inferences shifted from concrete to abstract over time (Study 1). The network models also indicated cultural differences in how one social inference was related to another between samples (Study 2, Asian N = 651, European N = 792). Together, these findings show that the high-dimensional network approach provides an alternative, non-mutually exclusive model for understanding the mental representation of social inferences in naturalistic contexts, which provides new insights into the dynamics and diversities of social inferences beyond the static, universal structure found with the low-dimensional latent construct approach.