RESUMO Este artigo discute uma investigação sobre as representações sociais (MOSCOVICI, 1978, 2003, 2007) de professores pré-serviço de inglês e espanhol sobre feedback corretivo oral. Participaram do estudo 64 estudantes concluintes de cursos de Letras das cinco diferentes regiões brasileiras. Uma vez que as representações sociais podem guiar os comportamentos, investigar as representações sociais dos participantes pode revelar como eles tendem a lidar com erros orais em suas futuras salas de aula. Os dados foram coletados por meio de um instrumento de associação livre de palavras e um grupo focal online. A análise revelou que os participantes ancoram feedback corretivo oral a um entendimento tradicional de correção do erro e aponta para a necessidade de incluir feedback corretivo oral no currículo de cursos de Letras.

ABSTRACT This article discusses an investigation about the social representations (MOSCOVICI, 1978, 2003, 2007) of pre-service teachers of English and Spanish on oral corrective feedback. The study population were 64 senior students of Letters programs from the five different Brazilian regions. Considering that social representations may guide behavior, researching the participants’ social representations may reveal how they tend to deal with oral errors in their future teaching. Data were collected using a free word association instrument and an online focus group. The analysis showed that the participants’ social representations are anchored on traditional understanding of error correction, and points to the need of including oral corrective feedback in the curriculum of Letters programs.

This paper is based on my master dissertation in which I investigated the washback effect of a university entrance exam on the teaching of the English language in four different high schools in Brasília. The methodology of the study, which focused on the practice of six English teachers, triangulates data from classroom observations, interviews with teachers and their responses to questionnaires, analyses of exam samples and the exam program. The results show that there is a slight washback effect on the course content and lesson planning, classroom materials and evaluation procedures. Evidence of washback effect is not found on either the language teaching approach or methodological procedures.