Page 47 - Getting of the fence
P. 47

                                three items (Cronbach α = .61). However, we deemed the content of this item of such importance that we decided not to eliminate this item from the scale.
2.3.1 Four approaches and one construct
In order to determine our understanding of a Comprehensive Approach to foreign language literature teaching and learning in which the four approaches are considered unified, we ran a Confirmatory Factor Analysis. CFA focuses on whether and the extent to which the four approaches are linked to the underlying latent trait (i.e. a Comprehensive Approach). Figure 2.2 shows the factor loadings of the four approaches regarding a Comprehensive Approach to foreign language literature teaching and learning.
Figure 2.2 Results of the CFA regarding a Comprehensive Approach to foreign language literature teaching and learning
Following the recommendations of Hu & Bentler (1999), the adequacy of model fit was evaluated on at least two statistics: a Comparative Fit Index (CFI) of >.95 and a Standardised Root Mean Square Residual (SRMR) of <.05 indicates a good fit. Furthermore, Tucker Lewis Index (TLI) value closest to 1.0 is also an indicator of a well-fitting model (Byrne, 2012). Results in Figure 2.2 show that the CFA resulted in a good fit of the model. The Text approach (.880) appears to have the strongest
Exploring EFL literature approaches
 coefficients in Table 2.6 show a range from relatively high (.88) to almost sufficient
(.61) for evidence of internal consistency for each of the four approaches. The
reliability analysis of the scale of the language approach showed that the Cronbach’s
α would be .64 if item ‘making reading miles to improve language skills’ would be 2 deleted, which is slightly higher than the reliability coefficient obtained with all
 45
 























































































   45   46   47   48   49