Application of Item Response Theory (IRT) in the identification of quantitative discriminating indicators for the practice of business benchmarking

Benchmarking as a management tool is based on practices and metrics. However, there are latent variables that cannot be measured directly. The Item Response Theory (IRT), a statistical technique that is widely explored in the health area, is little explored in the management area. This work seeks to...

Descripción completa

Detalles Bibliográficos
Autores: Baltazar, Marcos Charles Pinheiro, Albertin, Marcos Ronaldo, Pontes, Heráclito Lopes Jaguaribe
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2021
País:Brasil
Institución:Centro Universitário de Belo Horizonte (UNIBH)
Repositorio:Revista e-xacta
Idioma:portugués
OAI Identifier:oai:ojs.periodicos.uninove.br:article/16849
Acceso en línea:https://periodicos.uninove.br/exacta/article/view/16849
Access Level:acceso abierto
Palabra clave:Discrimination
Benchmarking index
Item Response Theory
Financial Performance Indicator.
Discriminação
Indicadores para benchmarking
Teoria de resposta ao item
Indicadores de desempenho financeiro.
Descripción
Sumario:Benchmarking as a management tool is based on practices and metrics. However, there are latent variables that cannot be measured directly. The Item Response Theory (IRT), a statistical technique that is widely explored in the health area, is little explored in the management area. This work seeks to apply the IRT to identify a set of indicators that are discriminative and possible to be used in the practice of business benchmarking. After analysing 31 indicators of 203 companies listed on the São Paulo Stock Exchange, it was possible to identify 21 discriminating indicators divided into 5 factors. It was concluded that IRT is a technique capable of identifying discriminatory indicators with low value of the Root Mean Square Error of Approximation (RMSEA) and that it brings new interpretations of already existing metrics - in the case economic and financial metrics - in addition to facilitating the understanding latent traits.