Assessment of evaluation instruments in learning processes

Abstract

The objective was to assess the instruments for evaluating the learning applied to students in the regular shift majors at URACCAN, Las Minas Campus, 2008. The assessment procedure was carried out through five stratified samples in the majors: Engineering, Administration, Social Sciences and Informatics for a level of significance of α = 0.05. Interviews and surveys were carried out with students and teachers, as well as a bibliographic review. It was found that nine categories of items are used in the evaluation instruments and that the planning is spontaneous. The most used evaluative instruments are short questions, short answers and true - false ones. The agreement of the evaluative instruments applied with the theories that support the pedagogical project is not exhaustive. There is little relationship between instructional aspects, number of reagents, representativeness of contents and cognitive levels evaluated and the distribution of the difficulty degree tends to be very easy. The most frequent type of item category was short questions (61%), the content validity assessed by these items were influenced by the level of depth required by the question. No formula was found to evaluate reliability for questions with generated response items, only for selected answers, but on the basis of the score assigned by the teacher to the evaluated assessment tools in Agroforestry Engineering, Business Administration Degree and Social Sciences with mention in Local Development and Animal Husbandry Engineering, alpha coefficients of 0.68, 0.48, -0.84 and 0.70 were obtained, respectively.