Aplicación del Teorema de Bayes en la selección de personal para disminuir la deserción laboral

Abstract

Bayes' theorem is an analytical method which does not rule out cases to calculate the probability of occurrence of an event, uses a statistic that can be continuously feedback and obtains better results as the number of cases to be considered increases. On the other part, job desertion is a natural phenomenon in companies, it occurs in different proportions according to variables such as gender, age, schooling or even according to the state of health of the people, which incite the unilateral voluntary disassociation of labor dependents. Conventional human talent management carries out specific evaluations in personnel selection processes, to link the company to people with the highest skills and abilities to carry out specific activities, however, it is not common for descriptive statistics to be used to support decision making at this stage. The present study identified eight profiles of job deserters according to the cases observed in Machala City through the use of descriptive statistics for the construction of a Bayesian network, at the beginning an exploratory investigation was carried out to recognize the problem, later it was theoretically based and described the variables associated with job desertion, then a survey was applied to a sample of 384 people to gather information that was processed through the rejection of cases, counting of frequencies and calculation of probabilities. The working situation of Machala shares some characteristics described in other studies, however, there were also anomalies regarding desertion according to gender and age of people. It is concluded that the Bayesian network of job desertion is a support tool for the personnel selection process which guides the choice of the ideal profile as the least prone to desertion.