Treinamento de professores para a realização de uma parte da análise de contingências: identificação da provável função do comportamento

Abstract

Many studies about the contingency analysis teaching for non-psychologists have been published. The procedures of some of these studies typically teach one of the necessary abilities to the contingency analysis achievement: the application of a previously defined standard procedure. Such studies generally don t make the apprentice to stay under control of what maintains the behavior wanted to be changed. Based on this, the present study intended to verify the efficacy of a training to teach teachers to raise a probable function of the behavior considered inadequate of the students in the classroom. Participated three elementary teachers, who had indicated one of their classrooms student that presented behaviors considered undesirable by them. In the baseline, eight written scenes of situations in the classroom that involved inadequate behaviors of students were presented to the participants. These behaviors could be maintained by positive reinforcement (attention gain from the teacher) or negative reinforcement (demand escape). The participants should answer, based only in the eight scenes, about the behaviors emitted by the students, the consequences, the antecedents, the frequencies that followed and the probable function of the behaviors. The training had 14 scenes. The experimenter provided analysis models for the first two scenes, answering the questions that accompanied the scenes. For every two scenes that followed, the answers for one of the analysis questions were removed, raising the number of questions to be answered by the participants. In test 1, equal to the baseline, the participants performance improved, almost reaching the maximum possible, which shows that the training was effective. A second test, the test 2, was made by presenting only the questions about what functions could have the behaviors presented in another eight scenes. In this test, two participants had their performance worse than in the test 1, what could indicate that they still needed the questions about the previous items and that it may be important to remove these questions gradually. A generalization test was made by asking the participants about the behavior probable function of the student from her classroom that had been indicated. Based on the experimenter s classroom observation, all the participants seemed to be able to identify the probable function of their student s behaviors, but when asked to propose an intervention based on the behavior function, only one of the participants did it, with partially adequate interventions, what shows that being able to identify the function of the behavior doesn t imply being able to propose an intervention