A Structural Equation Model on Job Stress and Productivity in Multi Sectors

Ganta Lavanya, Kanakam Deva Akshay, Gunna Srilekha, Boda Nikhil Sundar andDr Marada Srinivasa Rao
Department of Information Technology, MVGR College of Engineering (A), Vizianagaram, Andhra Pradesh,India

Abstract—Currently we are living in a cynical world where everybody wants to earn more. And for that, everybody runs more. Sometimes this extra running is at the expense of our health. It's not only physical but also mental that goes for a toss and Stress is the primary reason for unsatisfactory performance in today's dynamic workplace. Stress at work can result from a number of causes like workload, role ambiguity, job insecurity and the imbalance between work and family life. The study investigates the relationship between job stress and productivity by utilizing the Structural Equation Model (SEM) on data from four sectors namely IT, Hospital, Bank and Educational institutions. A questionnaire is used to collect data consisting of a variety of domains like stressor (job pressure, role stress, uppers, organization Structure), outcome (productivity, job performance) and some demographic variables. Factor analysis is used to validate the model through both Exploratory Factor Analysis (EFA) and Confirmatory Factor and Analysis (CFA). In addition, to improve the quality of the predictive model, machine learning techniques like Linear Regression, Random Forest, and Gradient Boosting Machine with cross-validation have been employed on stress and performance. The model will then help to identify the various factors that cause stress and on the basis of that, an organization can take necessary actions to reduce that stress and improve performance.Keywords: Job Stress, Productivity, Structural Equation Modelling, Machine Learning, Work-Life Balance, OrganizationalSupport.