In order to gather evidence on possible changes in student discourse a questionnaire was
included in tasks 1 and 10 (Gordon, 1998) (see Table 1). Students had to answer the
questionnaires individually. Questionnaires did not remain anonymous in order to identify
any change in discourse. This fact could cause a bias in the answers, given that students
could feel that they were being evaluated. For this reason, students were told that they would
obtain the maximum score assigned to the task simply for answering the questionnaire.
The final questionnaire (included in task 10) comprised questions structured in two blocks.
The first block includes the questions in the initial questionnaire. The second one includes
two kinds of questions: (a) those aiming to test the conclusions of previous studies; and (b)
those aiming to test the consistency of replies to enable interpretation of any change in
discourse