MULTINOMIAL REGRESSION FOR THE CHARACTERIZATION OF THE STRUCTURE OF ECUADORIAN COMPANIES
Keywords:
Multinomial Regression, business structure, multivariateAbstract
The objective of the research was to characterize by searching for the factors most related to the structure of the companies. For this, the questionnaire of the business structural survey ENESEM 2018 of the National Institute of Statistics and Censuses of a sample of 3723 medium and large companies, from the manufacturing, trade, service, mining and construction sectors of companies from all over Ecuador. Through a principal components analysis, a reduction of the groups of variables that made up the questionnaire was obtained, and through a multinomial model, the structure of the companies was characterized. Among the results obtained are the factors to characterize a large company in relation to characterizing it as a type A company, which is ultimately the variable "land", followed by the variable "mining". Likewise, the characteristic of "land" and “building” as well as the "mining and construction" sector also define more a company structured as large in relation to a medium-sized company B. It is concluded that large companies are characterized by the presence of Tangible Goods, while the medium type A and B by the other variables related to expenses, costs, production and depreciation. The model obtained was accepted for application by using the pseudo R2 Nagelkerke coefficient to test the suitability of the model, it was validated with an overall good classification rate of 87.3%.
Keywords: Multinomial Regression, business structure, multivariate.
Downloads
References
Acosta, R. J. H. (2016). Caracterización mediante análisis multivariane de los factores incidentes en la innovación de las empresas de servicios en el plan curricular de la educación superior colombiana. Encuentro Internacional de Educación en Ingeniería. Recuperado en: https://acofipapers.org/index.php/eiei/article/view/984
Almeida, J. G. B., Villegas, L. E. L., & Estéves, L. V. S. (2017). Estudio sobre la caracterización y fortalecimiento de los procesos comerciales de las MIPYMES de la Zona 1 del Ecuador. Comercio y Negocio, (7), 145-162.
Ortega, A. O. (2018). Enfoques de investigación. Extraído de https://www.researchgate.net/profile/Alfredo_Otero_Ortega/publication/326905435_ENFOQUES_DE_INVESTIGACION_TABLA_DE_CONTENIDO_Contenido/links/5b6b7f9992851ca650526dfd/ENFOQUES-DE-INVESTIGACION-TABLA-DECONTENIDO-Contenido.pdf
Fernández, V. P., & Fernández, R. S. M. (2004). Regresión logística multinomial. Cuadernos de la Sociedad Española de Ciencias Forestales, (18), 323-327.
Mero-Vélez, J. M. (2018). Empresa, administración y proceso administrativo. Revista Científica FIPCAEC (Fomento de la investigación y publicación en Ciencias Administrativas, Económicas y Contables). ISSN: 2588-090X. Polo de Capacitación, Investigación y Publicación (POCAIP), 3(8), 84-102.
INEC. (2019). Metodología de la ENESEM 2018. INEC, Dirección de Estadísticas Económicas - DECON, Quito. Recuperado el 07 de Mayo de 2020, de http://www.ecuadorencifras.gob.ec/documentos/webinec/Estadisticas_Economicas/Encuesta_Estructural_Empresarial/2017/2 017_ENESEM_Metodologia.pdf
Zamora, M. E. L., & Castro, L. M. U. (2021). Métodos estadísticos multivariantes aplicados en el estudio del rendimiento académico: Una revisión de la literatura. Revista Cognosis. ISSN 2588-0578, 6(1), 103-118.
