Modelo de riscos proporcionais: uma aplicação aos dados de reclamações de consumidores

Abstract

The usage of Statistical techniques, especially survival analysis, for data analysis has grown substantially in recent decades. This paper focus on the proportional risk model in which the study of the life time in which the response is the time until the occurrence of an event of interest is adjusted by covariables. Within the semiparametric models, the Cox model presents interesting flexibility and is widely used in survival data in which the effect of covariates of interest may vary over the duration of the study. In that sense, the objective here is to test and to evaluate the ability of the Cox model to accurately or satisfactorily model the time of conflict resolution using a tool through the Internet that allows direct interaction between consumers and companies. During the job, some difficulties have been faced and were resolved or circumvented. In the end, some regions, sex and some age groups were identified as relevants regarding the resolution of the complaints made in the portal Consumidor.com.br.