Effects on verbal and on nonverbal behavior of changes in reinforcement contingencies

Abstract

The present paper reports a study on the effects, both on verbal and nonverbal behavior, of changes on reinforcement contingencies applied to nonverbal behavior under contextual control. Such contingencies could be rules described by the experimenter, or  not, depending on the experimental condition. verbal behavior (descriptions of one’s own performance) was recorded periodically during each experimental condition, both before and after changes on reinforcement contingencies. Twenty children, from seven to eight years old, were exposed to procedures of contextual control of two conditional discriminations: oddity-from-sample-matching and identity matching-to-sample. The experiment had three phases: contextual control of reinforcement contingencies during Phase 1 were reversed during Phase 2, and reestablished during Phase 3. Participants were divided into three groups, according to the instructions presented during Phase1: Differential Reinforcement condition (RD) presented minimal instructions about how to proceed; Complete Instructions condition (CI) presented full instructions regarding the effective contingencies; Multiple Instructions condition (MI) presented three sets of instructions, accordingly to three changes on contextual control of reinforcement contingencies during Phase 1. For RD and CI participants, during Phase 1 a green light signaled identity matching and a red light signaled oddity contingencies, for MI participants the contextual control varied on Phase 1: initially green and yellow signaled, respectively, matching and oddity, then yellow and red, and finally gren and red (during Phases 2 and 3 only this third stimulus combination was employed): Transitions form one phase to the next were neither signaled nor instructed. During each phase participants were asked several times about what he/she should do in order to earn credits; this verbal behavior was never reinforced. Results showed that for all participants, in all phases, verbal behavior (self-descriptions) corresponded to nonverbal behavior (conditional discriminations performances). However, nonverbal behavior sensitivity to reinforcement contingencies varied: for RD participants nonverbal behavior was sensitive to reinforcement contingencies in all phases; while for CI and Mi participants it was not, instead it was controlled by the instructions presented by the experimenter at the beginning of the experiment.