Prof. Luís M. Grilo
University of Évora, Portugal

Title: Psychosocial risks to model employees’ stress: PLSc-SEM approach

Abstract: Structural Equation Modeling (SEM) is a powerful Multivariate Statistics technique that has recorded systematic scientific advances while having enormous interaction with many other research areas. The models presented here may eventually fall within the areas of Social, Behavioral and Health Sciences, since they involve some of the psychosocial risks to which workers in a Portuguese industrial company are subject to. The Copenhagen Psychosocial Questionnaire (COPSOQ), has been used internationally (Europe included) to assess the impact of these risks on workers’ health and well-being. Therefore, we used its medium version in a survey and based on the research hypotheses a reflective theoretical model was proposed, in which the endogenous latent variable “stress” (operationalized by manifest variables corresponding to the available questions in the COPSOQ) was the target variable of the study. Considering the full sample, the estimated model was obtained through the consistent Partial Least Squares (PLSc) estimator. The latent variables “quality of leadership” with effects on the mediating variables “justice” and “quantitative demands” and, in turn, “job satisfaction” with effect on “stress” revealed statistically significant. The “level of education” with two categories (with 9 levels of education as a break point) was also considered as a moderating variable, allowing to estimate two models. The path coefficient (representing a negative effect) between “job satisfaction” and “stress” was statistically significant only among workers with higher levels of education. These models estimated using PLSc-SEM provided some information considered relevant about the health and well-being of workers.

Bio: Luís Miguel Grilo (PhD in Mathematics and Statistics, University of Lisbon, 2006) is Statistics Professor in the Department of Mathematics at the University of Évora (UÉ), Portugal. He is currently an integrated member of CIMA - UÉ (Research Center in Mathematics and Applications) and collaborating member of NOVA Math, NOVA University of Lisbon, Portugal, where he has been developing scientific research in Distribution Theory (exact and near exact distributions of some statistics used in Multivariate Analysis) and working on statistical data analysis and statistical modeling, with a special interest in Engineering, Social and Health Sciences. He has collaborated on some (inter)national projects, with applications of Structural Equation Modeling to assess burnout in students and workers. From 2014 to 2020 he was a member of the board of CLAD (Portuguese Association of Classification and Data Analysis). He was editor of some international journals, such as the special issue “Advances in Computational Data Analysis” (2020), of the Journal of Applied Statistics (JAS). He is currently a special issue editor on WCDANM2024 in JAS and also an Associate Editor for the Research in Statistics Journal, Taylor & Francis. As a member of Scientific and Organizing Committees, he has participated in several (inter)national Mathematics and Statistics meetings. In particular, he has been the local chair of some international conferences and the chairman of the executive committee of the Workshop on Computational Data Analysis and Numerical Methods (WCDANM).