Construction of the scale of quality perceived by electric power services offered by Brazilian companies

The global demand for electricity and the assessment of service quality provided by electricity utility companies requires robust methodological approaches to ensure accurate and reliable measurements. Traditional methods, such as Classical Test Theory (CTT), have been applied to the evaluation of latent traits, but they present limitations concerning reliability. Alternative methods derived from Psychometrics, such as Confirmatory Factor Analysis (CFA) and more recently, Item Response Theory (IRT), overcome these limitations. IRT offers advantages over CTT as it takes into account the difficulty and discrimination of each item, allowing for a more accurate estimation of the latent trait. In contrast, CTT uses only the total score, which can lead to inaccurate assessments. This paper applies Item Response Theory (IRT) to develop and validate a perceived quality of services scale. Using data from the Consumer Satisfaction Questionnaire of the Brazilian Electricity Regulatory Agency (ANEEL), Confirmatory Factor Analysis (CFA) and IRT validated the scale’s unidimensional structure, demonstrating high reliability, as well as the difficulty and discrimination of the items. The results identified four levels of perceived quality, with most consumers classified as satisfied or very satisfied, although regional disparities were observed—satisfaction was highest in the South and lowest in the North. These findings reinforce the advantages of IRT in service quality assessment, enabling detailed response pattern modeling and the development of standardized, interpretable scales. The validated instrument provides regulators and utility companies with a valuable tool for monitoring service quality and optimizing policies to enhance consumer satisfaction. Future research may refine benchmarking methodologies for cross-market comparisons.