Abstract: Superstitious behaviors, as exceedingly common phenomenon, take place frequently in our daily life, which can have either negative or positively effect on people's life. In addition, they can jeopardize moral education in colleges and schools. For example, the superstitious phenomenon may make people rather vain about their prediction or even cause people to kill others in fear. Therefore, it's extremely essential to investigate the mechanism of the occurrence of superstitious behaviors. There are two major psychological theories which can explain the mechanism of the occurrence of superstitious behaviors. They are behavioral sect and cognitive sect. From the perspective of behavioral sect, Skinner proposed the accidental reinforcer-response pairings account, and after that, other psychologists addressed control illusion, a vital concept in the superstitious field, and presented control heuristic theory and regulatory focus explanation from the cognitive perspective. However, there are drawbacks in both explanations, so we need a more integrated account for the procedure of superstitious behaviors. Based on two experiments, this research examines the differential effects of non-contingent positive and negative reinforcement operations on the acquisition of superstitious behaviors. In the first experiment, 72 participants were randomly and evenly divided into four groups, including the positive group (6s), positive group (3min), the negative group(6s) and negative group(3min). All participants were instructed to try to produce and/or keep the positive pictures (all from IAPS) on a computer screen (positive reinforcement), or try to prevent and/or remove the negative pictures from the screen (negative reinforcement) by using response keys. In the second experiment music experimental material (all were pre-tested) was used, and the stimuli schedule was different from the first experiment while other conditions remained the same. For instance, all 80 participants were randomly and evenly separated in four groups: the positive group (6s), positive group (3min), the negative group(6s) and negative group(3min). All participants were instructed to try to produce and/or keep the positive music on a computer audio amplifier (positive reinforcement), or try to prevent and/or remove the negative music from the audio amplifier (negative reinforcement) by using response keys. Data from the two experiments indicated that participants exposed to lean schedules of negative reinforcers believed that they had greater control over non-contingent stimulus events than participants exposed to either rich or lean schedules of positive reinforcers. These findings suggest that, compared to positive reinforcement, negative reinforcement operations may provide a more fertile condition for the development and maintenance of superstitious behaviors. And compared to rich schedule, lean schedule seems to make participants believe there was more control than rich schedule under the negative reinforcement operations, but the data show the opposite variation direction of illusion of control under the positive reinforcement operations. These conclusions will help us to predict the future tendency of the occurrence of superstitious behaviors more accurately. Meantime, this can help us to adjust the moral educational programs for students in colleges and schools.