Mapeamento de repertórios de leitura e escrita em escolas com baixos índices na Prova Brasil

Abstract

The Brazilian Ministry of Education is effectively measuring text comprehension by the national evaluation called Prova Brasil. However, lowered averages performance recurrent propagation for the schools raises two assumptions: the first one, that students may not have fully learned previous repertoires, more basic, which might influence their performance in text comprehension measures; and the second one, is related to the fact that Prova Brasil presents an informative character, not having as main objective to provide specific teaching procedures linked to the measures used. This Prova Brasil feature make arduous for the teachers and coordinators to program interventions that produces improvements in this complex repertoire. Both studies of this work aimed to contribute with the discussion of these assumptions. In Study 1 basic reading and writing repertoires were characterized regarding regular and irregular isolated words for 5th grade students of three schools with low rates in Prova Brasil. The results indicated deficits for varied basic repertoires in the three schools, however, were observed most affected performances for writing under dictation and for irregular words. In Study 2, it was described the items construction that composed a reading comprehension assessment, based on a behavioral interpretation. Evidences for validity and reliability were also investigated by the items, using Item Response Theory (IRT). The results not only corroborate the low median performance presented at Prova Brasil by the students, but also allowed verify the validity and consistency of the items to evaluate intermediate reading comprehension repertoire. The two studies allowed mapping reading and writing repertoires, from basic to complex, showing the feasibility of the two evaluations to prescribe specific teaching procedures in order to remediate the deficits identified.