Desenvolvimento de métodos experimentais para investigar relações causais no comportamento "supersticioso"

Abstract

Skinner (1948) carried out an experiment demonstrating that "temporal accidents" between behavior and reinforcer were sufficient to modify the response. Such patterns of behavior have been similarly compared to "human superstitious" behavior. Although an interesting analogy, it often obscures the questions that these data raise about the notions of causality that underlie behavioral science. The present study investigated the literature to find ways to develop experiments that explore the notions of cause that permeate "superstitious" behavior. Experiment 1 investigated the proximity of the notions of "superstitious" behavior, operant behavior and contingency. Experiment 2 investigated an alternative way of using pairing time to learn "superstitious" behavior. Experiment 3 investigated the possibility of learning "superstitious" behavior through function transfer. The results of Experiment 1 left some simple distinctions between operant and "superstitious" even more fragile, the results of Experiment 2 demonstrate some problems in the experimental control of the previous experiment and Experiment 3 demonstrated that it's possible to establish a behavior similar to the "superstitious" behavior without the use of non-contingent reinforcement. The results of the three experiments demonstrated that the concept of "superstitious" behavior causes difficulties and limitations in the experimental investigation, as previously noted in the theoretical review. As a solution, the paper proposes the substitution of the term "superstitious" behavior by notions of causality