Estudio psicométrico de los estudiantes de la carrera de Estadística en la asignatura de estadística inferencial, periodo 2021-2022

Abstract

The objective of this work was to determine the contents of low, medium and high academic performance, through an Inferential Statistics test in the Statistics degree students of the ESPOCH, period 2021 - 2022. For the test development, building stages given by the National Institute for the Evaluation of Education were used, considering 22 topics taught in the analytical program of the Inferential Statistics subject, with a total of sixty-six questions. To perform the analysis of the quality of the test, Psychometrics was used applying two theories, on one hand with the Classical Test Theory the reliability, standard error of the mean, degree of difficulty and the discrimination of the items were studied, on the other hand with the Response Theory to the test item the logistic models of one, two and three parameters were studied, comparing and analyzing the values presented when a new parameter is added. A frequency analysis allowed classifying students according to their academic level. Once the test was done, it was determined that it presents a good quality in the phases of design, construction, application, calibration and a poor quality in the verification stage, since a pilot study was not conducted. In the content analysis, four sub-topics were detected where there is low academic performance, eighteen of medium level, and no sub-topic has high-level students. The students showcase major problems in the contrasting hypotheses topic. It is recommended to update the test taking the best items as input.