In the ADHD group, we observed a significant difference in the effect of age on ACC between BM-Local and BM-General (standardised β 5 =0.462, p<0.001) and marginally significant differences in the effect of age on ACC between BM-Local and BM-Global (standardised β 4 =0.228, p=0.073).
Second, as I currently understand the reported model results, the fact that ß4 (representing the difference in relationship to SRS scores between task 2 and task 1; the authors refer to ß3 here although I assume they mean ß4) is negative and shows a trend towards significance would actually mean the relationship between BM processing accuracy and SRS scores is more negative for task 2 relative to task 1 and not, as the authors state, that the correlation with SRS scores is greater for task 1.
The values of ß4 (representing the difference in relationship to SRS scores between BM-Global and BM-Local) and ß5 (representing the difference in relationship to SRS scores between BM-General and BM-Local) were positive and showed a trend towards significance, indicating that the correlations with SRS total score were more negative for BM-Local relative to BM-Global and BM-General: “A general linear model was constructed (Table S2, Model 4): SRS = β0 + β1 * ACC + β2 * D1 + β3 * D2 + β4 * (ACC * D1) + β5 * (ACC * D2).