Distribución geométrica censurada aplicada a datos de postulantes a la Universidad Nacional Agraria La Molina.

Abstract

This research deals with the geometric distribution with randomly censored data. The data set, provided by the Universidad Nacional Agraria La Molina (UNALM), corresponds to applicants during the years 2018 to 2021, for whom it was recorded whether they achieved admission or not. Data was randomly censored, and it was considered that a censorship was presented when the applicant did not enter and did not apply again. The estimated parameters were the probability that the applicant would enter and the probability that he would give up his attempts, which would be the probability of censorship. Both quantities were estimated with the maximum likelihood method and the confidence intervals were obtained from the Fisher information matrix. Thus, it was found that the probability of entering the university is 0,117, and this is higher in men. In addition, a probability of censorship was estimated as 0,296.