Fig. 1

Density plots based on the models applied in the Bayesian framework. The plots depict the distributions of the posterior mean estimations of the three coefficients corresponding to the components of affective dysregulation with regard to distress and valence. Distributions of the respective coefficient estimates (coefficient intercept corresponding to the subcomponent affective homebase in a, coefficient within-variance corresponding to the subcomponent within-person variability in b, and  coefficient autocorrelation corresponding to the subcomponent attractor strength in c) are shown for each investigated group separately for dataset 1 (patients with borderline personality disorder [BPD], those with posttraumatic stress disorder [PTSD], those with bulimia nervosa [BN] and healthy controls [HC]) and dataset 2 (patients with borderline personality disorder [BPD], those with panic disorder [PD], those with major depression [MD] and healthy controls [HC])