Composite quantile estimation in PLS-SEM for environment sustainable development evaluation

The main purpose of our article is to provide statistical methods and quantitative evidences regarding environment sustainable development evaluation (ESDE). To accomplish our investigation, we establish a theoretical ESDE model with environment and its factors first, and then develop a composite-quantile-based partial least square algorithm (CQ-based PLS) in a modified structural equation model (SEM) with quantiles. The real data analysis proofs the hypotheses in our theoretical ESDE model and provides quantitative estimates of both path and loading coefficients in CQ-based ESDE model. Taking the ordinary PLS and the existing quantile-based PLS algorithms as references, we illustrate the statistical performances of CQ-based PLS-SEM estimators through bootstraps. Our investigations mainly illustrate that environment development has positive impacts on human resource, education and health while effects economy negatively. Economy is positively affected by human resource and health. Education positively impacts economy, human resource and health. Compared with the existing two PLS algorithms, CQ-based PLS-SEM estimators have relatively better performances on the whole, which increases the possibility for generalization of our model and algorithm in various applications.