University students’ intentions to learn artificial intelligence: the roles of supportive environments and expectancy–value beliefs

Table 4 Test for invariance of path coefficients across field, gender, and year

Models	χ²	df	CFI	TLI	RMSEA	SRMR	Δχ²	Δdf	p
Field of study
Unconstrained model	750.02	352	0.921	0.914	0.070	0.064	11.39	8	0.18
Constrained model	761.41	360	0.920	0.916	0.070	0.077	11.39	8	0.18
Gender
Unconstrained model	879.89	352	0.902	0.894	0.078	0.063	10.35	8	0.24
Constrained model	890.24	360	0.901	0.896	0.077	0.073	10.35	8	0.24
Year of study
Unconstrained model	793.77	352	0.919	0.912	0.071	0.066	16.28	8	0.04
Constrained model	810.05	360	0.917	0.913	0.071	0.069	16.28	8	0.04

Path coefficients in the unconstrained model were free to vary across groups, while path coefficients in the constrained model were constrained to equality across groups