Avaliação nacional do rendimento escolar: um novo modelo de apresentação dos resultados por escola utilizando descritores

Abstract

This research aims to contribute to the improvement of one of the fragile dimensions of the large-scale evaluation system that exist in Brazil. Namely, the feedback of the Prova Brasil to the public directly involved in it such as teachers, students and the management and pedagogical team. According to this perspective, this project associates with the theoretical frameworks that challenge the vertical accountability of schools, their agents and their harmful effects of pressure on the magisterial entity and the competitive logic. In this sense, this theory accords with the view about the evaluation system proposed by Freitas (2016). Thus, it is proposed a new model of presenting the results of any evaluation that will allow a review of the institution that is being tested by the institution itself, through specific and significant information about the subjects involved in the process. This new report model considers the relevance of the schools in obtaining understandable results on the performance of their students in the Prova Brasil and connecting them with studies that confirm the lessen impact of this data in the learning process (VIANNA, 2003) or the alienation of it in the complex process of interpreting the scores (SILVA, 2010). In order to reach the goal of this project, a broad study with the contribution of the documentary sources inventory and the bibliographic review of the topic on the Prova Brasil are necessary. It is possible to know the different stages the test had, its process of institutionalization and the characteristics of its different periods through the inventory while the bibliographic review allowed us to understand the main approaches as well as the mapping of the scientific productions concerning the presentation, dissemination and results of this evaluation process. In subsequent stages, the research highlights the components of Prova Brasil and the elements associated with the production and publication of its results per school, as well as the descriptions, information and analysis on the elements that guide the tests. In order to do so it is necessary an in-depth understanding about the main concepts and theories that support the current evaluation model created by INEP, among which, the Item Response Theory (IRT); the concept of the Matrix Reference and Descriptors, Scales of Proficiency and Proficiency Levels that are referenced in the works of Andrade, Tavares and Valle (2000). These contribute to the understanding of IRT and the perspective of comparability, to the detriment of the specific results of schools. In addition, during the creation of the methodological course of the new report model per school, this research portrays the microdata base, which is its main source, and examines its possibilities over processing information related to the proposed report model. Due to this, the project is guided by important references related to the Database different areas (DATE, 2000), to the Electronic Spreadsheets (COTTINGHAM, 1999) and to Statistics (TRIOLA, 1998). Finally, it is concluded that the model proposed here, which is complementary to that of the INEP, contributes to bringing the results of evaluations closer to educational institutions and their community. Thus, it meets different objectives of the report made available by INEP. This is possible because it does not aim at a temporal comparison, at a ranking position or at a government goal achievement. Rather, it is concerned with the result.