3.1.Favoritism scale confirmatory factor analysis (CFA)
Structural Equation Modeling (SEM) may be viewed as a general model of
many commonly employed statistical models, such as analysis of variance,
analysis of covariance, multiple regression, factor analysis, path
analysis, econometric models ofsimultaneous equation and nonrecursive
modeling, multilevel modeling, and latent growth curve modeling. Readers
are referred to Tabachnick & Fidell (2007) for an overview of many of
these methods. Through appropriatealgebraic manipulations, any one of
these models can be expressed as astructural equation model. Hence, SEM
can be viewed as an “umbrella” encompassing a set of multivariate
statistical approaches to empirical data, both conventional and recently
developed approaches.(Jichuan; and Xiaoqian, 2020)
The first-level factorial structure of the Favoritism in Ship crew
management scale consisting of three dimensions and a total of 22 items
(5 items in operational processes favoritism, 8 items in social rights
favoritism and opportunities and 7 items in demographic and personal
characteristics favoritism) was tested by using the AMOS 21 software. As
a result of the first-level confirmatory factor analysis (CFA) of the
scale, 2 items were kept out of the model due to errors, and the
analysis was repeated. The goodness of fit values obtained as a result
of the first-level CFA are given below.
Table 5. Structural equation modeling conformity values