Table 1.

Goodness of fit statistics for different models

Variable		Model	χ2	df	P	AIC
N240	Neutral	ADE	1.90	3	.59	−4.09
		AE	2.19	4	.70	−5.82
	Fear	ACE	4.74	3	.19	−1.26
		AE	5.13	4	.27	−2.87
	Happy	ADE	.24	3	.97	−5.76
		AE	.43	4	.98	−7.57
P3	Neutral	ADE	6.04	3	.11	.04
		AE	6.14	4	.19	−1.85
	Fear	ADE	2.47	3	.48	−3.53
		AE	2.79	4	.59	−5.21
	Happy	ADE	6.83	3	.08	.83
		AE	7.19	4	.13	−.81
P600 (identity)		ADE	2.75	3	.43	−3.25
		AE	2.95	4	.57	−5.05

χ² shows the goodness of fit, with lower values indicating a better fit; AIC is Akaike’s Information Criterion. Significance levels: * P < 0.05; ** P < 0.01. The difference in goodness-of-fit between the full (ADE or ACE) model and reduced (AE) model $\mathrm{\Delta }{\chi }_{(\mathit{\text{df}}=1)}^2={\chi }_{\text{ADE}}^2-{\chi }_{\text{AE}}^2$ did not reach significance for any of the variables, indicating that D or C paths can be dropped without significant deterioration of fit