Page 139 - Hojnik, Jana. 2017. In Persuit of Eco-innovation. Drivers and Consequences of Eco-innovation at Firm Level. Koper: University of Primorska Press
P. 139
Methodology 139
tions, collectively referred to as a model. A model can include relations
among measured variables and latent variables (i.e., factors, constructs)
as well as nondirectional and directional (direct and indirect) relations.
The model typically is presented at two levels: conceptual and statisti-
cal (Hoyle and Panter 1995). The conceptual model specifies the relations
among concepts that are operationalized in the empirical study, while
the precise statistical model that will be tested cannot be deduced from
the presentation of the conceptual model. As such, each construct repre-
sented in the conceptual model must be operationalized, and the mod-
el must be translated into the statistical manifestation that has been or is
to be tested. A path diagram can be an effective means of communicat-
ing structural equation models at the statistical level (Hoyle and Panter
1995). Also, in our study, a structural equation model was used to test the
theoretical model. SEM is a statistical methodology that takes a confirm-
atory (i.e., hypothesis-testing) approach to the multivariate analysis of a
structural theory bearing on some phenomenon (Byrne 2006 in Murovec
et al. 2012). Typically, this theory represents “causal” processes that gen-
erate observations on multiple variables (Bentler 1995).
SEM is a family of statistical models that seek to explain the relation-
ships among multiple variables (Hair et al. 2009). It therefore examines
the structure of interrelationships expressed in a series of equations, sim-
ilar to a series of multiple regression equations. These equations depict
all of the relationships among constructs (the dependent and independ-
ent variables) involved in the analysis. Constructs are unobservable or la-
tent factors represented by multiple variables (much like variables repre-
senting a factor in factor analysis). So far, each multivariate technique has
been classified either as an interdependence or a dependence technique.
SEM can be thought of as a unique combination of both types of tech-
niques because SEM’s foundation lies in two familiar multivariate tech-
niques: factor analysis and multiple regression analysis (Hair et al. 2009).
Hair et al. (2009) emphasized three main characteristics based on
which SEM models can be distinguished:
- Estimation of multiple and interrelated dependence relation-
ships,
- An ability to represent unobserved concepts in these relation-
ships and correct for measurement error in the estimation pro-
cess,
- Defining a model to explain the entire set of relationships.
tions, collectively referred to as a model. A model can include relations
among measured variables and latent variables (i.e., factors, constructs)
as well as nondirectional and directional (direct and indirect) relations.
The model typically is presented at two levels: conceptual and statisti-
cal (Hoyle and Panter 1995). The conceptual model specifies the relations
among concepts that are operationalized in the empirical study, while
the precise statistical model that will be tested cannot be deduced from
the presentation of the conceptual model. As such, each construct repre-
sented in the conceptual model must be operationalized, and the mod-
el must be translated into the statistical manifestation that has been or is
to be tested. A path diagram can be an effective means of communicat-
ing structural equation models at the statistical level (Hoyle and Panter
1995). Also, in our study, a structural equation model was used to test the
theoretical model. SEM is a statistical methodology that takes a confirm-
atory (i.e., hypothesis-testing) approach to the multivariate analysis of a
structural theory bearing on some phenomenon (Byrne 2006 in Murovec
et al. 2012). Typically, this theory represents “causal” processes that gen-
erate observations on multiple variables (Bentler 1995).
SEM is a family of statistical models that seek to explain the relation-
ships among multiple variables (Hair et al. 2009). It therefore examines
the structure of interrelationships expressed in a series of equations, sim-
ilar to a series of multiple regression equations. These equations depict
all of the relationships among constructs (the dependent and independ-
ent variables) involved in the analysis. Constructs are unobservable or la-
tent factors represented by multiple variables (much like variables repre-
senting a factor in factor analysis). So far, each multivariate technique has
been classified either as an interdependence or a dependence technique.
SEM can be thought of as a unique combination of both types of tech-
niques because SEM’s foundation lies in two familiar multivariate tech-
niques: factor analysis and multiple regression analysis (Hair et al. 2009).
Hair et al. (2009) emphasized three main characteristics based on
which SEM models can be distinguished:
- Estimation of multiple and interrelated dependence relation-
ships,
- An ability to represent unobserved concepts in these relation-
ships and correct for measurement error in the estimation pro-
cess,
- Defining a model to explain the entire set of relationships.
