We examined the choices of 52 adolescents (18 years of age); 26 were enrolled in an alternative to incarceration program and 26 were university undergraduates. Respondents were presented with a gambling task and were shown two spinners, one offering a sure gain (or loss), the second offering a gamble (probabilities of the gamble included ?, 1/3 and ?; magnitudes of the gamble included low, medium and high monetary amounts). Participants received 18 trials, nine framed as gains and nine framed as losses (each gain condition was analogous to a loss condition) and individuals were asked to report their confidence after each scenario. In the gain frame, respondents began with nothing and were offered a choice between a sure gain and a gamble offering the possibility of winning larger amounts or winning nothing. For the loss frame, individuals were given an endowment and could lose an amount for sure or take a chance and risk losing nothing or risk losing everything. In each corresponding gain and loss pair, the net sum that could be won was kept equal. In addition, respondents completed surveys on sensation seeking, behavioral inhibition and behavioral activation.
Data analysis was conducted both without and with the confidence ratings. In both regressions, delinquents scored significantly higher on sensation seeking in the gain frame, but not in the loss frame. In the first ANOVA, magnitude was a significant effect as was the interaction between magnitude and the population of the participant. Delinquents were more likely to choose the gamble regardless of the magnitude, while undergraduates picked the sure option at higher magnitudes. Also, a three way interaction was found between framing, probability and magnitude. The second ANOVA, replicated these findings and probability became a significant effect. Such results highlight the potential benefit of programs that work to enhance decision-making processes in order to prevent initial crime or potential recidivism.