Page 198 - Hojnik, Jana. 2017. In Persuit of Eco-innovation. Drivers and Consequences of Eco-innovation at Firm Level. Koper: University of Primorska Press
P. 198
In Pursuit of Eco-innovation
The model with three factors gave goodness-of-fit indexes (NFI = 0.928;
NNFI = 0.945; CFI = 0.954; SRMR = 0.044; RMSEA = 0.086), and the
Cronbach’s alpha was 0.952. The standardized coefficients were all pos-
itive, high (over 0.50) and statistically significant. Correlation between
product and process eco-innovation was estimated at 0.79, correlation be-
tween product and organizational eco-innovation was 0.65, and the pro-
cess and organizational eco-innovation dimensions also showed high cor-
relation (0.68). All correlations were statistically significant. We can see
that confirmatory factor analysis demonstrated better goodness-of-fit in-
dexes for the model with three dimensions than for the model with two
dimensions. Therefore, we decided to use the three-factor model solution.
Furthermore, statistical information of each eco-innovation dimen-
sion’s internal consistency (Cronbach’s alpha reliability) and convergence
198 (goodness-of-fit model indexes) based on the overall sample (N = 223) is
indicated in Table 56. Table 56 summarizes the statistics for all eco-inno-
vation dimensions (product, process and organizational eco-innovation)
and further illustrates, for each eco-innovation dimension, model fit in-
dexes, range of standardized coefficients, Cronbach’s alpha reliability and
the number of items included. We can see that Cronbach’s alpha is high
in all cases – for the product, process and organizational eco-innovation
dimensions separately as well as for the eco-innovation construct (over
0.80). More specifically, the dimension of product eco-innovation showed
good reliability (Cronbach’s alpha = 0.872) and convergence in terms of
coefficients. The other two dimensions, process eco-innovation (Cron-
bach’s alpha = 0.912) and organizational eco-innovation (Cronbach’s al-
pha = 0.956), showed excellent reliability and convergence in terms of co-
efficients. Moreover, standardized coefficients are all positive, high (over
0.50) and statistically significant. The goodness-of-fit indexes are also
high; only RMSEA values for process and organizational eco-innovation
showed slightly worse fit. Moreover, the goodness-of-fit indexes are bet-
ter when related to the entire eco-innovation construct. Lastly, the mod-
el showed goodness-of-fit indexes (NFI = 0.928; NNFI = 0.945; CFI =
0.954; SRMR = 0.044; RMSEA = 0.086). We can see that the goodness-
of-fit indexes are better with a three-dimension model of eco-innovation,
which is in line with our theory. However, the exploratory factor analysis
gave two dimensions as solutions (joining product and process eco-inno-
vation dimensions), probably also because of highly related dimensions;
product and process eco-innovation demonstrated high correlation (r =
0.79; see Table 58). Thus, we decided on three dimensions of eco-innova-
tion on the basis of the results of the confirmatory factor analysis.
The model with three factors gave goodness-of-fit indexes (NFI = 0.928;
NNFI = 0.945; CFI = 0.954; SRMR = 0.044; RMSEA = 0.086), and the
Cronbach’s alpha was 0.952. The standardized coefficients were all pos-
itive, high (over 0.50) and statistically significant. Correlation between
product and process eco-innovation was estimated at 0.79, correlation be-
tween product and organizational eco-innovation was 0.65, and the pro-
cess and organizational eco-innovation dimensions also showed high cor-
relation (0.68). All correlations were statistically significant. We can see
that confirmatory factor analysis demonstrated better goodness-of-fit in-
dexes for the model with three dimensions than for the model with two
dimensions. Therefore, we decided to use the three-factor model solution.
Furthermore, statistical information of each eco-innovation dimen-
sion’s internal consistency (Cronbach’s alpha reliability) and convergence
198 (goodness-of-fit model indexes) based on the overall sample (N = 223) is
indicated in Table 56. Table 56 summarizes the statistics for all eco-inno-
vation dimensions (product, process and organizational eco-innovation)
and further illustrates, for each eco-innovation dimension, model fit in-
dexes, range of standardized coefficients, Cronbach’s alpha reliability and
the number of items included. We can see that Cronbach’s alpha is high
in all cases – for the product, process and organizational eco-innovation
dimensions separately as well as for the eco-innovation construct (over
0.80). More specifically, the dimension of product eco-innovation showed
good reliability (Cronbach’s alpha = 0.872) and convergence in terms of
coefficients. The other two dimensions, process eco-innovation (Cron-
bach’s alpha = 0.912) and organizational eco-innovation (Cronbach’s al-
pha = 0.956), showed excellent reliability and convergence in terms of co-
efficients. Moreover, standardized coefficients are all positive, high (over
0.50) and statistically significant. The goodness-of-fit indexes are also
high; only RMSEA values for process and organizational eco-innovation
showed slightly worse fit. Moreover, the goodness-of-fit indexes are bet-
ter when related to the entire eco-innovation construct. Lastly, the mod-
el showed goodness-of-fit indexes (NFI = 0.928; NNFI = 0.945; CFI =
0.954; SRMR = 0.044; RMSEA = 0.086). We can see that the goodness-
of-fit indexes are better with a three-dimension model of eco-innovation,
which is in line with our theory. However, the exploratory factor analysis
gave two dimensions as solutions (joining product and process eco-inno-
vation dimensions), probably also because of highly related dimensions;
product and process eco-innovation demonstrated high correlation (r =
0.79; see Table 58). Thus, we decided on three dimensions of eco-innova-
tion on the basis of the results of the confirmatory factor analysis.
