Table 2.

Results of confirmatory factor analysis of study variables.

Model	χ 2	df	Δχ 2	Δdf	CFI	TLI	RMSEA	SRMR
1.6-factor	877.69	476	–	–	0.92	0.91	0.06	0.06
2.5-factor	1,063.11	481	185.42	5	0.88	0.87	0.07	0.06
3.4-factor	1,146.98	485	83.87	4	0.80	0.78	0.09	0.08
4.3-factor	1,254.19	489	107.21	4	0.67	0.64	0.13	0.11
5.1-factor	1,642.19	492	388.00	3	0.55	0.50	0.15	0.12

df, degrees of freedom; Δχ², chi-square differences; Δdf, degrees of freedom differences; CFI, comparative fit index; TLI, Tucker-Lewis index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual.

N = 247.

Model 1 (six-factor model) includes all study variables. Model 2 (five-factor model) combines on-the-job embeddedness and off-the-job embeddedness as one factor and considers the other four variables as four independent factors. Model 3 (four-factor model) combines on-the-job embeddedness and off-the-job embeddedness as one factor, promotion focus and prevention focus as one factor, and considers the other two variables as two independent factors. Model 4 (three-factor model) combines on-the-job embeddedness and off-the-job embeddedness as one factor, promotion focus and prevention focus as one factor, and career adaptability and proactive career behavior as one factor. Model 5 (one-factor model) combines all six variables as one factor.

Δχ² and Δdf reflect differences of χ² and df between the corresponding model and Model 1. All χ² and Δχ² are significant at the p < 0.001 level.