Page 137 - Hojnik, Jana. 2017. In Persuit of Eco-innovation. Drivers and Consequences of Eco-innovation at Firm Level. Koper: University of Primorska Press
P. 137
Methodology 137
ed exploratory and confirmatory factor analysis for all constructs used in
the eco-innovation model using two statistical packages, SPSS and EQS
6.1. Furthermore, to test the hypotheses pertaining to the influence of
eco-innovation antecedents (environmental policy instruments, mana-
gerial environmental concern, customer demand, expected benefits and
competitive pressure) on eco-innovation (product, process and organiza-
tional eco-innovation and eco-innovation construct) and its consequenc-
es (company performance (in terms of growth and profitability), econom-
ic performance, competitive benefits and internationalization), we used
the multivariate technique of structural equation modeling (hereinafter
SEM) employing the statistical program EQS 6.1. Therefore, the model
and hypotheses were tested by using SEM, which allows for simultaneous
evaluation of multiple related dependent and independent relationships
and takes into account measurement error (estimates) in the evaluation
process (Hair et al. 1998).
Most scales used in this study were examined for convergent and
discriminant validity using exploratory and confirmatory factor analy-
ses. For each construct used in this study, exploratory factor analysis has
been performed. We have therefore tested whether the number of factors
proposed by the exploratory factors analysis is in line with the expected
number of factors. We used the Maximum Likelihood method and Di-
rect Oblimin rotation (oblique rotation, which expects correlations be-
tween factors). After conducting the exploratory factors analysis, we have
also conducted confirmatory factor analysis for each construct to assess
the reliability, validity and goodness-of-fit of each construct. Confirma-
tory factor analysis (CFA) enables us to test how well the measured var-
iables represent the constructs (Hair et al. 2009). Regarding the eco-in-
novation construct, which was measured as a second-order construct,
we first checked for construct reliability, which measures the reliabili-
ty and internal consistency of the measured variables representing a la-
tent construct (Hair et al. 2009). Before assessing the construct validi-
ty which deals with the accuracy of measurement (the extent to which a
set of measured variables actually reflects the theoretical latent construct
those items are designed to measure), we have to establish construct reli-
ability (Hair et al. 2009). After this, we checked the eco-innovation con-
struct, which includes three dimensions for convergent validity (the ex-
tent to which indicators of a specific construct converge or share a high
proportion of variance in common) and discriminant validity (the extent
to which a construct is truly distinct from other constructs). There are
several ways to estimate the relative amount of convergent validity among
ed exploratory and confirmatory factor analysis for all constructs used in
the eco-innovation model using two statistical packages, SPSS and EQS
6.1. Furthermore, to test the hypotheses pertaining to the influence of
eco-innovation antecedents (environmental policy instruments, mana-
gerial environmental concern, customer demand, expected benefits and
competitive pressure) on eco-innovation (product, process and organiza-
tional eco-innovation and eco-innovation construct) and its consequenc-
es (company performance (in terms of growth and profitability), econom-
ic performance, competitive benefits and internationalization), we used
the multivariate technique of structural equation modeling (hereinafter
SEM) employing the statistical program EQS 6.1. Therefore, the model
and hypotheses were tested by using SEM, which allows for simultaneous
evaluation of multiple related dependent and independent relationships
and takes into account measurement error (estimates) in the evaluation
process (Hair et al. 1998).
Most scales used in this study were examined for convergent and
discriminant validity using exploratory and confirmatory factor analy-
ses. For each construct used in this study, exploratory factor analysis has
been performed. We have therefore tested whether the number of factors
proposed by the exploratory factors analysis is in line with the expected
number of factors. We used the Maximum Likelihood method and Di-
rect Oblimin rotation (oblique rotation, which expects correlations be-
tween factors). After conducting the exploratory factors analysis, we have
also conducted confirmatory factor analysis for each construct to assess
the reliability, validity and goodness-of-fit of each construct. Confirma-
tory factor analysis (CFA) enables us to test how well the measured var-
iables represent the constructs (Hair et al. 2009). Regarding the eco-in-
novation construct, which was measured as a second-order construct,
we first checked for construct reliability, which measures the reliabili-
ty and internal consistency of the measured variables representing a la-
tent construct (Hair et al. 2009). Before assessing the construct validi-
ty which deals with the accuracy of measurement (the extent to which a
set of measured variables actually reflects the theoretical latent construct
those items are designed to measure), we have to establish construct reli-
ability (Hair et al. 2009). After this, we checked the eco-innovation con-
struct, which includes three dimensions for convergent validity (the ex-
tent to which indicators of a specific construct converge or share a high
proportion of variance in common) and discriminant validity (the extent
to which a construct is truly distinct from other constructs). There are
several ways to estimate the relative amount of convergent validity among
