Figure 1.

Diagram of multi-level structural equation models used for analysis

Note. The left panel shows the latent decomposition of observed variables y into within person (y_it) and between-person (y_i) variance for individual i during interaction t. The right panel shows the model used for all analyses. Single-headed arrows represent regression paths, double headed arrows represent variances. Circles indicate latent variables and squares are observed variables. Note that the predictor, Personality Pathology_i, is shown as an observed variable but in the general personality pathology models it is a latent variable. ℇy_i reflects individual differences in within-person variance, and y_i is the estimate of an individual’s average. 𝜁y_i represents variance in y_i not accounted for by predictors in the model.