Integrated structural equation modeling and causal steps in evaluating the role of the mediating variable

Validation of causal inference data using DirectLiNGAM in an environmental small-scale model and calculation settings

“…Considering causal inference methods, it is important to check their calculation methods and evaluate their results [ 19 , 20 ]. In this paper, we confirmed the basic usage of DirectLiNGAM and validated the result of causal inference with DirectLiNGAM.…”

“…In addition, as validations for confirming credibility, we validated the correlation coefficients between factors, feature importance for the target variable, intervention effect and optimal value for the effect, and investigation with propensity score as effective methods even for small-scale data. Although various methods of path analysis have been reported [ 13 , 20 , 21 ], DirectLiNGAM provides a method with new point of view.…”

Implementation of assertive communication through leadership style and its implications for organisational commitment

“…The method used in this research is Regression Analysis with Causal Step Test proposed by Baron & Kenny (2016) using SPSS Version 26 software. According to Baron & Kenny in (Syafiq et al, 2022) mention that to determine the significance of the intervening variable in mediating the relationship between the dependent and the independent factor, the analysis was carried out with a causal steps approach. This method suggests a set of requirements that must be fulfilled for a mediation model (Nurofik et al, 2022) The Baron and Kenny Causal Test using SPSS involves several calculation steps.…”

Social integration and utilization of national basic public health services among China's internal migrants with chronic diseases: A structural equation modelling approach

“…Structural equation modelling (SEM) is an appropriate statistical method for multidimensional variables and has been widely used in recent years. Specifically, SEM is a statistical method for analyzing the relationship between variables based on a covariance matrix that merges factor analysis and path analysis methods to simultaneously examine the interaction between multiple variables, including the direct and indirect effects between variables [ 29 , 30 ]. Moreover, SEM compensates for the fact that traditional statistical analysis methods cannot measure multivariate problems directly and accurately.…”