Exploração de capacidades de compreensão: o que os livros didáticos ensinam e o que as provas avaliam?

Abstract

A common problem to teachers and students dissatisfied with results obtained in tests points out to difficulties in reading questions. Teachers complain that their students cannot read; on the other hand, students claim that they cannot answer some specific questions becausethey do not understand what is asked of them. Is the problem related to the student, who fails to understand the objective of the question? Do questions require capabilities that have not been developed by the students? Or is the problem related to the differences in theformulation of questions: i.e. the student learns to deal with a certain kind of question formulation and in the test there are different kinds of questions? In the present work, one regards the principle that the teaching of reading has the textbook as the only teaching tool.Having this in mind, one formulates two hypotheses; the first is that there is a difference between the teaching of reading developed by means of the textbook and the evaluation of reading in tests. Therefore, students learn in a manner and are evaluated in another. As a second hypothesis, one believes that the questions in tests often leave gaps that makecomprehension more difficult. The corpus of the present research is composed of the questions of reading activities present in the Portuguese Language book of the collection Português Linguagens, 7th year of Primary School, whose authors are William Roberto Cerejaand Thereza Cochar Magalhães, edition published in 2006. The corpus is also composed of questions selected from six examples of Portuguese Language tests taken by 7th-year students of a same class, from February up to November 2010. In this work, one also considerslanguage as an interaction activity and reading as a cognitive, social and cultural process to build up meaning. The analysis of the corpus of this research is mainly based on Bakhtins works and on Text Linguistics researchers such as DellIsola (2007), Kleiman (2008), Solé(1998), Marcuschi (2008), Cafiero (2005), Moretto (2008), Hoffmann (2005). The results found in this research support the hypotheses previously made. There is a difference regarding the contextualization of texts and the genre approach presented in textbooks and tests.Concerning the design of questions in tests, there is a prevalence of multiple choice questions that privilege negative statements and questions without contextualization with the generic use of imperative verbs that hinder comprehension.