Figure 7.
Models based on self-report and neural data. These diagrams show seven models predicting individual control-averse behavior (y), based on self-reports of social cognition (S), freedom restoration (F), negative affect (A), and subjectwise estimates of right IPL–dlPFC connectivity in the controlled minus the free condition (PPI). Arrows indicate main effects and the line with a circular endpoint in model 10 indicates an interaction effect. The bar graphs show the BIC and R2 for each model, with the winning model highlighted in black.