Agrupamento de questões de formulários do SAEB para obtenção de fatores que podem interferir no processo de aprendizagem

Abstract

Annually, in Brazil, questionnaires are administered to the school community at the time of the test conducted through the Basic Education Assessment System. This is a large-scale external assessment system, consisting of a set of instruments carried out periodically by the National Institute for Educational Studies and Research Anísio Teixeira. This assessment aims to measure the quality of education in both public and private schools, serving as a crucial tool in basic education in Brazil. However, the data resulting from these questionnaires are complex and voluminous, necessitating the use of Data Science to summarize the data in order to facilitate the extraction of valuable information. Thus, this work proposes to cluster questions from the forms used in this assessment to obtain factors that may interfere with the learning process. For this purpose, question groups are identified through the application of factor analysis to represent specific characteristics related to schools. In this work, questionnaires applied to 9th-grade students and teachers, as well as school managers of the investigated schools, are analyzed. These responses are explored through graphs and summary measures. As a result, a better visualization of the information contained in the questionnaires can be obtained, along with the presentation of a metric for the respondents’ perception of the identified factors. As a future proposal, the metric values for respondents’ perceptions can be used in a multilevel data regression model to measure the impact of these values on students’ performance in the test administered when the questionnaires are applied. Furthermore, the proposed methodology can be applied in different contexts and with different questionnaires, thereby increasing its scope and usefulness.