Modelo basado en ingeniería kansei para el diseño de ambientes laborales en call centers de Bogotá y su área metropolitana

Abstract

This monograph aims to formulate a model for the design of work environments in call centers located in Bogotá and its Metropolitan Area, utilizing Kansei Engineering for its conceptualization, and tools such as neural networks and structural equations for validation. The approach seeks to provide guidance for designers in creating more efficient, productive, and satisfying workspaces for employees. To formulate the model, a preliminary scientometric review was conducted to identify relevant studies, trends, characteristic factors, and dimensions for the design process. Subsequently, both the frequency of appearance in studies and the results of the initial application of Quality Function Deployment (QFD) were considered to prioritize key factors and dimensions for design. Once these key dimensions were identified, a set of adjectives describing them was created to guide the formulation of questionnaire questions on a semantic differential scale. Through a pilot test and expert input, the questions were refined, revised, and perfected. With the instrument validated, data collection took place using specialized social networks such as LinkedIn and Call Center groups with members matching the worker profile in the Bogotá-Metropolitan Area. After data collection, a detailed statistical analysis was performed to appropriately characterize the studied sample. Subsequently, to reduce the number of factors and simplify the analysis, SPSS software and linear regression were employed to construct preliminary equations to understand the relationship between positive and negative emotions, the work environment, and the four levels of the consumer needs pyramid, along with the generated key factors. An attempt was made to validate the regression results using structural equation modeling through IBM AMOS software. However, it was found that the studied variables exhibited linear dependence and their numerical values showed no significant differences, hindering the use of this method. As an alternative, the multilayer perceptron was chosen as a neural network model to validate the relationships between variables and constructs, determining the importance of emotional dimensions in workplace design and the feasibility of applying Kansei Engineering in this process. Additionally, a recommended procedure was presented to achieve the design of healthy work environments.