Contribuições para o controle experimental na aprendizagem de discriminações por bebês.

Abstract

The literature has emphasized the necessary investigation of the variables acting in conditional discrimination learning for infants in experimental and natural situations. Part of the challenge in reaching the research goals with this population lies in the difficulties of dealing with young children during experimental situations, controlling the presence of the adult who is the source of social reinforcements and in the necessity of considering the peculiarities in the population s behavior. The work objective was to identify efficient experimental variables to teach simple discrimination and identity matching tasks to infants. Three studies were conducted. In Studies 1 and 2, data was collected from four 12 to 25-month-old participants in an experimental situation organized at their daycare center. Several toys exerted different stimulus functions and were exposed to the infants in an apparatus especially designed for this study. The participants worked on simple discrimination and identity matching tasks, during which the following variables were manipulated: training procedures; play characteristics with the adult during the moments of manipulation of toys other than the experimental ones; play characteristics with the adult while using the toys that had the Sr role. The decisions about the manipulation of variables were made based on constant analysis of each participant s performance. In the third study, data was collected from one 17-month-old participant in his own house. Pictures of animals classified as familiar or unfamiliar to the infant exerted stimulus function and were presented by the mother in matching auditory-visual tasks, matching dictated names and figures. The three studies results showed how efficient the manipulation of variables is to simple and conditional discrimination learning examination. The variables considered critical were related to designed experimental conditions, including the training procedures, the stimulus selection, the gauging of responses and the reinforcement nature. The general discussion emphasized the beneficial relation between experimental research and natural situation investigations, due to the possibility of identifying stimulus classes, response classes and appropriate procedures to teaching simple discrimination and identity matching to infants.