Fig. 4: Chain mediation results.
From: The neurocognitive mechanism underlying math avoidance among math anxious people
These mediation models examine the mechanisms linking math anxiety to math avoidance. In each model, the independent variable (X) represents math anxiety scores, and the dependent variable (Y) reflects the level of math avoidance. The first mediator (M1) captures individual sensitivity to task difficulty in predicting avoidance behavior, while the second mediator (M2) represents the activation strength of a specific brain region during decision-making. The numbers next to the arrows indicate regression coefficients in the mediation models, and asterisks denote significance levels (*p < 0.05, **p < 0.01, ***p < 0.001). a Mediation model with hippocampal (HPC) activation as M2. b Mediation model with posterior insula (PI) activation as M2. c Mediation model with mid-cingulate cortex (MCC) activation as M2. *p < 0.05, ** p < 0.01, *** p < 0.001.
