The perception of control in loud noise

Perception of control is known to affect performance under stress. Two experiments are reported the object of which was to find out how loud noise during a contingency assessment task influences perceived control. Subjects were required to choose one of two responses, note one of two results, and then provide an overall percentage estimation of the degree of contingency present after 40 trials. Subjects made these judgments for one of three levels of objective contingency (25%, 50%, 75%), either in quiet (55 dBA) or in loud noise (95 dBA) conditions. The first experiment involved a series of randomly chosen, preprogrammed outcomes for noncontingent trials. An unexpected effect of noise was that subjects improved their successes in predicting events, and could only have done so by finding sequential structure in the preprogrammed alternations. They also overestimated control relative to contingency data actually received, at the 25% objective contingency level, but the result could have been dependent on different base levels of data actually received. A second experiment, with a random generation of outcomes for noncontingent trials, resulted in no differences in success levels, but confirmed that noise is associated with the overestimation of contingency at the 25% objective contingency level and demonstrated the same effect for the 50% level. The results are discussed in the context of the 'illusion of control'.