Three-stage Bibliometrics-structural Equation Modelling–Artificial Neural Networks to Predict Green Production in Business: Global Research Impact on Pro-environmental Planned Behaviour

Authors:
Publisher website
Google Scholar
Abstract
Goods and services are being shifted to be greener in terms of their production. This study investigated the determinant factors that influence pro-environmental green production businesses. An assessment of pro-environmental planned behaviour was used to determine green production. The data were taken from global research studies worldwide based on Scopus indexing to be analyzed with bibliometrics. The collected data consisted of 186 on perceived environmental authority support, 5,501 on perceived environmental concern, 3,755 on pro-environmental attitude (PEATT), 396 on pro-environmental subjective norms, 477 on pro-environmental perceived behavioural control and 407 on green production intention. The co-occurrence network was taken as the input for the partial least squares structural equation model to be formed into six factors. The measurement follows the inner and outer model analyses. The latent variables generated were used as an input for a multilayer perceptron artificial neural network (ANN) with 10 neurons. The bibliometric results show the range of documents from 1975. The structural equation model shows that 5 out of 10 hypotheses are accepted. ANNs predict the normalized importance of each determinant factor, where PEATT is explored as the strongest and most important predictor. Several insights, such as providing incentives, subsidies, rewards, promotion, education, socialization and a deep understanding of benefits for green production, are strongly suggested.

This publication has 66 references indexed in Scilit: