Effects of the experimental history with different kinds of instructions and the control of contingencies for following instructions

Abstract

Instructional control – the control of the listener’s behavior by the verbal behavior of others, has a major role in learning process, including those responsible for the maintenance of cultural practices through generations. Despite its importance, many questions about behaviors governed verbally remain open to scientific investigation. For example, some studies have pointed out that instructed behaviors can be insensitive to their consequences while others have suggested that insensitivity could be a product of specific conditions involved in experimental histories. The present study investigated the effects of experimental histories with different kinds of instructions (correspondent to or discrepant from contingencies), preceded or not by a history of contingency shaped behavior, on the behavior of following correspondent instructions and on the maintenance of instructional control when contingencies were changed without any warning. Twenty-eight undergraduate students, designed to two groups, were exposed to a matching-to-sample task, either identity or oddity matching-to-sample, under contextual control, across two experimental sessions. The first sesión established the experimental history. During this session, participants of Group I were initially exposed to a minimal instruction in order to establish control by contingencies; participants of Group II were not.  Additionally, half of the participants in each group (I and II) were exposed to two 20-trial blocks that began with the presentation of a correspondent instruction; the other half was exposed to discrepant instructions. During the second session, tests were carried out. In these tests, the behavior was established by a correspondent instruction and, after that, contingencies were reversed without any further instructions. During the test session all participants were exposed to the following sequence of four 20-trial blocks: 1) novel correspondent instructions; 2) contingencies reversed (without any warning); 3) novel correspondent instructions; 4) contingencies reversed (without any warning). In the experimental history session, all participants of Group I presented control by contingencies; half of the participants in Group II followed the corresponding instructions; participants exposed to instructions discrepant from contingencies split, in the sense that some continued following discrepant instructions, while others abandoned the instruction-following behavior. During tests, in the second session, 13 out of 14 participants from Group I showed behavioral sensitivity to changes in contingencies, that is, their behavior was in accordance with correspondent instructions but they abandoned this pattern of responding under discrepant instructions, in early stages of each testing block. In Group II, 5 out of 14 participants showed similar results. There was no difference in sensitivity considering the kind of instruction to which participants were exposed in the first session. These results give support to previous studies suggesting a conceptual revision in the interpretation of instructed behavior, considering that insensitivity should not be included as a property of instructed behavior, but should be seen as a dimensión that varies as a function of a series of variables, including the participants´ previous experience.