Eyewitness accuracy and confidence: within- versus between-subjects correlations

Previous researchers using between-subjects comparisons have found eyewitness confidence and accuracy to be only negligibly correlated. In this study, we examined the predictive power of confidence in within-subject terms. Ninety-six subjects answered, and made confidence ratings for, a series of questions about a crime they witnessed. The average between-subjects and within-subject accuracy-confidence correlations were comparably low: r = .14 (p less than .001) and r = .17 (p less than .001), respectively. Confidence is neither a useful predictor of the accuracy of a particular witness nor of the accuracy of particular statements made by the same witness. Another possible predictor of accuracy, response latency, correlated only negligibly with accuracy (r = -.09 within subjects), but more strongly with confidence (r = -.27 within subjects). This pattern was obtained for both between-subjects and within-subject comparisons. The theoretical and practical implications of these results are discussed.     

Instructions and reinforcement in the observing behavior of adults and children

Observing responses of college students and children of different ages (4 to 5, 9 to 10, and 13 to 14 years) were studied using two-component schedules of reinforcement. Different frequencies or numbers of points were given in the components and points were backed by either money or instructions about getting points (e.g., “points are for the experimenter's information only”). Stimuli that were either correlated or uncorrelated with the components were contingent on presses on either of two concurrently available levers (observing responses). In each of three experiments, preferences for stimuli were consistent with the conditioned-reinforcement hypothesis of observing but inconsistent with the uncertainty-reduction hypothesis. Stimulus preferences were established, eliminated, or reversed by instructions about the significance of getting points not backed with money. Instructions to in effect ignore points with no monetary value eliminated stimulus preference provided that money was not associated with points in other conditions for the same subjects. Stimulus preferences were otherwise independent of monetary point value. Instructions that described reinforcement contingencies promoted more rapid adjustment to the contingencies than occurred without them, and preferences were less variable across sessions within subjects and were on the average more extreme. Although some results involving absolute rates of observing depended upon the age of subjects, stimulus preferences did not. The data suggest that the conditioned-reinforcement hypothesis of observing has considerable generality for humans in choice procedures in the laboratory. That monetary value of points did not affect preferences in a graded fashion is consistent with versions of the conditioned-reinforcement hypothesis which emphasize reduction in delay to reinforcement over other reinforcement parameters.     

Complex decision making: A longitudinal study of process and performance

Two studies using a computer-simulated, strategy-formulation game and business students were conducted using simultaneous verbal protocols. It was found that a number of information-processing and information-evaluation thought processes were significantly related to game performance. Consistent with what is known about individual decision making, the present results suggest that, for the task used in this study, individuals who engage in causal analysis perform better than those who do not. But those who focus on negative emotions, blindly repeat previously successful decisions, and engage in illogical through processes perform more poorly than those who do not. Causal replication using real managers within organizational settings, however, is needed before firm conclusions can be drawn from this research.