As a post hoc analysis, IQ was added into the multi-group model. This model showed acceptable fit to the data (Chi square[526]=692.76; p<.001; Chi square/df ratio=1.32; CFI=.93; RMSEA=.04). Among H/I children, BL IQ was significantly associated with Neuropsychological functioning at F1 (β =.60; p<.001) but not with ADHD severity (β =−.09; p=.35). As a result of inclusion of IQ, Neuropsychological functioning at F1 was not significantly associated with ADHD severity at F2 (β =−.17; p=.10) and SES was no longer significantly associated with ADHD or NEPSY. Other paths remained substantially similar to the model without IQ. Among TD children too, IQ was significantly associated with F1 Neuropsychological functioning (β=.61; p<.001) and marginally associated with ADHD severity (β= −.23; p=.06). Other paths did not change substantially.
