Health Research Center, University of Almería.
Psicothema. 2020 May;32(2):221-228. doi: 10.7334/psicothema2019.275.
This study offers a comparative analysis of evidence for the predictive validity of SAVRY and YLS/CMI scores in predicting risk of recidivism in a group of young people who received a Juvenile Justice order.
The sample was made up of 594 youths aged between 14 and 18 (M=15.63, SD=1.08) at the time they committed an offense.
Both instruments showed high accuracy in predicting recidivism, with the greatest accuracy observed in the SAVRY and YLS/CMI total scores, as well as in the Individual domain of the SAVRY. Comparative analysis of the AUCs of both instruments indicated no statistically significant differences between total scores from the two instruments. Results showed statistically significant differences in comparisons of means and AUCs between the groups of young reoffenders and non-reoffenders in all cases. Our results did not support the hypothesis that dynamic risk factors are a better predictors of recidivism in young offenders.
This study offers empirical evidence of the predictive capacity and differential functioning of the SAVRY and YLS/CMI instruments in the Spanish context.
本研究对 SAVRY 和 YLS/CMI 评分在预测一组接受少年司法命令的年轻人的累犯风险方面的预测有效性进行了比较分析。
该样本由 594 名年龄在 14 至 18 岁之间（M=15.63，SD=1.08）的青少年组成，他们在犯罪时。
两种工具在预测累犯方面都具有很高的准确性，SAVRY 和 YLS/CMI 总分以及 SAVRY 的个体领域观察到的准确性最高。对两种工具的 AUC 进行比较分析表明，两种工具的总分之间没有统计学上的显著差异。结果表明，在所有情况下，在重新犯罪的年轻罪犯和非罪犯群体之间，均值和 AUC 的比较都存在统计学上的显著差异。我们的结果不支持动态风险因素是年轻罪犯累犯更好预测因素的假设。
本研究提供了 SAVRY 和 YLS/CMI 工具在西班牙背景下的预测能力和差异功能的实证证据。
