The importance of familial risk factors in children with ADHD: direct and indirect effects of family adversity, parental psychopathology and parenting practices on externalizing symptoms

Table 2 Model fit parameters for CFA, SEM 1, SEM 2, and the alternative model

Model	χ² (df)	p	CFI	SRMR	RMSEA	AIC	BIC	Δχ² (df)	p
CFA	5.21 (9)	0.82	1.00	0.01	< 0.001
SEM 1	46.85 (24)	0.003	0.98	0.03	0.04
SEM 2	32.51 (23)	0.09	0.99	0.02	0.03	13,570.21	13,751.60	14.16 (1)^a	< 0.001
Alternative model	110.27 (23)	< 0.001	0.93	0.08	0.09	13,655.55	13,836.94

In the CFA model the validity of the measurement models of the three latent factors parental psychopathology (pPSYC), child ADHD (cADHD), and child ODD (cODD) was assessed. In SEM 1, direct and indirect effects of family adversity (FAI), parental psychopathology (pPSYC), positive parenting (pPAR), and negative parenting (nPAR) on child ADHD and ODD symptoms were examined (FAI → pPSYC → pPAR/nPAR → cADHD/cODD). In SEM 2, SEM 1 was extended to include the error covariance of positive and negative parenting. The alternative model contained the following alternative arrangement of the familial factors, with otherwise unchanged paths: pPSYC → FAI → pPAR/nPAR → cADHD/cODD
AIC = Akaike information criterion, BIC = Bayesian information criterion, CFA = confirmatory factor analysis, CFI = comparative fix index, RMSEA = root mean square error of approximation, SRMR = standardized root mean square residual
^areference model = SEM 1

ISSN: 1753-2000