Resumen

HERRERO-CORONA, Luis. Predictive model for the selection of a public opinion measurement technique. The Anáhuac j. [online]. 2021, vol.21, n.2, pp.50-77.  Epub 14-Mar-2022. ISSN 2683-2690.  https://doi.org/10.36105/theanahuacjour.2021.v21n2.02.

The objective of this article is to build a predictive model, based on information from projects carried out in market research and public opinion survey companies, choosing the recommended data collection method, which can be face to face interviews, telephone or online surveys, according to the requirements of each case.

The predictive model type is one of classification, and several are built and analyzed using decision tree data mining techniques, discriminant analysis, K nearest neighbor analysis, and neural network analysis. Additionally, a segmentation of contacts in clusters is carried out to complement and enrich the knowledge provided by the classification techniques.

It is concluded that the models generated by both decision trees and neural networks are the ones that best predict the public opinion measurement technique to be used.

Palabras llave : surveys; public opinion; predictive model; C38; C45; C83; M30; M31.

        · resumen en Español     · texto en Español     · Español ( pdf )