Table 3.

Model comparison, goodness of fit indices.

ModelCM/dfNFITLICFIRMSEA
Good Model Fit<3> .95> .95> .95< .07
Acceptable ModelFit<5> .90> .90> .90.07 -.10
1. One-factor model (12 items)2.34.80.81.87.11
2. One-factor model (11 items)3.64.71.66.76.12
3. One-factor model (10 items)3.38.75.77.81.14
4. Two-factor model (9 items)1.57.91.94.96.07
5. Two-factor model (Tsui and Kennedy, 2009; 12 items)2.07.82.85.90.10
6. Two-factor model (Cheung, 2006; 12 items)2.24.81.82.88.10
7. Two-factor model (Cheung, 2006; 11 items)2.23.90.83.84.10
8. Three-factor model (Tschannen and Hoy, 2001; 12 items)2.09.83.84.90.10
Assessing the fitness of the 2-factor structure by Tsui and Kennedy (2009) was accomplished with marginal fit. The goodness-of-fit indexes included: χ(53)= 109.45, p < .001; CMIN/DF= 2.07, CFI= .90, NFI = .82, TLI= .85 and RMSEA= .10 (90% CI of RMSEA= .07-.12). The factor loadings for efficacy in classroom management (Items 1, 2, 6, 7, 8) ranged from .54 to .77 and for efficacy for teaching and support (Items 3, 4, 5, 9, 10, 11, 12), from .58 to .77. The two models (Cheung, 2006; Tsui and Kennedy, 2009) were almost identical. The only difference between these two models was that Item 2 was moved from the teaching and support factor to the classroom management factor.
>As the assumption of normality was not violated, and no outliners were found, the means and standard deviations of the three factors were obtained. A series of one sample t-tests or ANOVAs found no significant differences between the two factors for the three independent variables (Table 4).
[ViewLarge]
This Feature Is Available To Subscribers Only

or Create an Account

Close subscription notice
Close access options