Acceso usuarios
Pedro José Horcajo Gil, Víctor Dujo López, José Manuel Andreu Rodríguez, Marta Marín
Se conceptualizan las distintas aproximaciones teóricas y metodológicas sobre valoración y gestión del riesgo de reincidencia delictiva en menores infractores. Se realiza una revisión sistemática de los instrumentos empleados con mayor frecuencia para valorar y gestionar el riesgo de reincidencia. Se emplean las bases de datos bibliográficos PsycINFO y Google Académico. Se seleccionan aquellos instrumentos cuyas puntuaciones presentan adecuados niveles de validez predictiva y aquellos cuyas características resultan relevantes por su idiosincrasia. Se localizan quince instrumentos: tres escalas actuariales (JSORRAT-II, Static-99 y PCL:YV) y doce instrumentos clínico-estructurados (YLS/CMI, SAVRY, SIED-AJ, EARL-20B, EARL-21G, J-SOAP-II, ERASOR 2.0, MEGA, Asset, ARMIDILO-S, DASH-13 y PREVI-A). Existe una gran variabilidad en los resultados en cuanto a validez predictiva cuando se comparan diferentes estudios realizados con un mismo instrumento. La adecuación de un instrumento al contexto judicial debe basarse en resultados de valores de validez predictiva (AUC) a partir de .70-.75.
Both theoretical and methodological approaches were gathered based on the risk assessment and management of criminal recidivism among young offenders. A systematic review of the instruments most often used to assess and manage the risk of recidivism was carried out. PsycINFO and Google Scholar were used as databases. The instruments chosen were based on adequate levels of predictive validity and those whose aspects were relevant. Fifteen instruments were found: three actuarial scales –JSORRAT-II, Static-99, and PCL-YV– and twelve structured clinical judgment measures –YLS/CMI, SAVRY, SIED-AJ, EARL-20B, EARL-21G, J-SOAP-II, ERASOR 2.0, MEGA, Asset, ARMIDILO-S, DASH13, and PREVI-A. There is a great variability in the results found in terms of predictive validity when different studies with the same tool are compared. The adequacy of an instrument to the judicial context must be based on values of predictive validity (AUC) between .70-.75.
