A B S T R A C TIn Study 1, eyewitness identification of the perpetrator of a "crime" (fire), framed as either intended or unintended, was studied in 138 children, ages 7 to 18. Analysis using Signal Detection reveals an interaction of age and condition on decisional bias. Like in past studies, the framing of the act had no effect on the 7-to 9-year-olds, but did have an effect on decisional bias for the other age groups. Decisional bias was more lax (indicting more false alarms) in the intended condition for 10-to 12-and 14-to 15-year-olds but was more stringent (fewer false alarms) for the 16-to 18-year-olds. This pattern of age and condition differs from the pattern of explicit judgments (how bad the act was, how much punishment it deserved, and how bad it is to commit a false alarm or a miss). Study 2 was conducted to confirm the unexpected findings for the 10-to 12-year-olds. Forty-two children, ages 10 -12, viewed the same film, which was framed as unintended but resulting either in (a) major or (b) minor damage (fire). Approximately half randomly assigned to condition (a) and half to (b). Parallel results were obtained with an earlier study, with lower bias scores (more false alarms) in the major than minor damage conditions. Thus, from both studies, we may conclude that decisional bias is more lenient (resulting in more false alarms) for 10-to 12-year-olds when either intent or damage is bad.
S C I E N T I F I C A B S T R A C TThese studies are based on the assumption that when adults, adolescents, or children identify someone as the "guilty" one (the person who committed the act), they are not only making an identification based on memory and thinking, but also a moral decision. This is because, by the act of identifying or not identifying someone, the eyewitness runs the risk of either convicting an innocent person (making a false-positive error) or letting a guilty person go free (a false-negative error). Our interest is less in the overall accuracy of their identifications and more in the balance of false-positive and false-negative errors. We have found in these and past studies that the balance of these 2 kinds of errors changes with age, and that this pattern may also depend on (a) the child's general understanding of the purpose of the task, which appears to be "lost" on 7-to 9-year-olds, the youngest group studied, and (b) for older children and adolescents, how the act is described; for example, intended or not. In this way, we can understand that the act of identifying the perpetrator as a moral decision and not simply an act of perception and memory.