Page 203 - Hojnik, Jana. 2017. In Persuit of Eco-innovation. Drivers and Consequences of Eco-innovation at Firm Level. Koper: University of Primorska Press
P. 203
Results
tion dimensions-only model) is based on the assumption of the non-un-
idimensionality of the eco-innovation concept. These two models are
nested in the model with both the dimensions and the common factor,
a method that allows for model comparisons (Antončič 2002; Ruzzier
2005). These comparisons are shown in Table 59.
Table 59: The dimensions-only vs. the one common factor model
Chi-square df NFI NNFI CFI SRMR RMSEA
920.656 89
M1: One common *** 87 0.712 0.682 0.731 0.109 0.205
factor model
228.463 68 0.928 0.945 0.954 0.044 0.086
M2: Dimen- ***
sions-only model 21 203
(three dimensions) 142.048 19
*** 0.956 0.963 0.976 0.033 0.070
M3: Model with
both the dimen- 778.608 0.846
sions and the com- *** 0.378
mon factor
86.415
M1-M3 ***
M2-M3
Note: Chi-square: * significant at p < 0.05; ** significant at p < 0.01; *** significant at p <
0.0001.
The one common factor model indicated an overall poor fit relative
to the dimensions-only model in all goodness-of-fit indexes. Model fit in-
dexes of the dimensions-only model and the model with both the dimen-
sions and the common factor are very high. The model with both the di-
mensions and the common factor has somewhat lower residuals (SRMS)
and errors (RMSEA) and higher NFI, NNFI and CFI indexes.
The contributions of the two models are shown in the last two rows
of Table 59. Both Chi-square differences are significant (p < 0.0001), in-
dicating that both models may contribute to explanatory power. How-
ever, the NFI for the two model differences, computed with the formula
from Bentler (1990) by including models 1 and 3 respectively, demon-
strates that the contribution of the dimensions seems to be quite sub-
stantial (NFI = 0.846), while the contribution of the overall-factor model
seems to be relatively minimal (NFI = 0.378). Overall, the one common
factor model seems to be inferior to the dimensions-only model. This can
tion dimensions-only model) is based on the assumption of the non-un-
idimensionality of the eco-innovation concept. These two models are
nested in the model with both the dimensions and the common factor,
a method that allows for model comparisons (Antončič 2002; Ruzzier
2005). These comparisons are shown in Table 59.
Table 59: The dimensions-only vs. the one common factor model
Chi-square df NFI NNFI CFI SRMR RMSEA
920.656 89
M1: One common *** 87 0.712 0.682 0.731 0.109 0.205
factor model
228.463 68 0.928 0.945 0.954 0.044 0.086
M2: Dimen- ***
sions-only model 21 203
(three dimensions) 142.048 19
*** 0.956 0.963 0.976 0.033 0.070
M3: Model with
both the dimen- 778.608 0.846
sions and the com- *** 0.378
mon factor
86.415
M1-M3 ***
M2-M3
Note: Chi-square: * significant at p < 0.05; ** significant at p < 0.01; *** significant at p <
0.0001.
The one common factor model indicated an overall poor fit relative
to the dimensions-only model in all goodness-of-fit indexes. Model fit in-
dexes of the dimensions-only model and the model with both the dimen-
sions and the common factor are very high. The model with both the di-
mensions and the common factor has somewhat lower residuals (SRMS)
and errors (RMSEA) and higher NFI, NNFI and CFI indexes.
The contributions of the two models are shown in the last two rows
of Table 59. Both Chi-square differences are significant (p < 0.0001), in-
dicating that both models may contribute to explanatory power. How-
ever, the NFI for the two model differences, computed with the formula
from Bentler (1990) by including models 1 and 3 respectively, demon-
strates that the contribution of the dimensions seems to be quite sub-
stantial (NFI = 0.846), while the contribution of the overall-factor model
seems to be relatively minimal (NFI = 0.378). Overall, the one common
factor model seems to be inferior to the dimensions-only model. This can
