Development of the Youth Version of the CARE
Kathryn Seifert, Ph.D.
The CARE (Child & Adolescent Risk Evaluation) is a screening tool that assesses behavioral problems in children and teens (Seifert, Phillips, and Parker, 2001). This assessment tool measures both risk and protective factors for behavioral disorders. The risk categories for both instruments are individual characteristics and history (such as history of violence, enuresis, poor anger management, and psychosis), peer interactions (for example, bullying behaviors and deviant peer group), educational, school, and work problems (like learning problems, lack of work success, and school failure), and family characteristics and dynamics (such as a history of harsh disciplinary practices and family violence during childhood). There is one resiliency category (for example, supportive adult and school or work success). Static and dynamic factors are used, as well. These fourty-nine items are those factors that have been identified in the research literature as being associated with severe behavior problems and aggression (Seifert, 2000; Quinsey, Harris, Rice, & Cormier, 1998). The Youth CARE is based on the idea that the more risk factors and the fewer resiliency factors that a person has, the greater his/her risk for violent recidivism or other behavior problems.
There are 825 individuals in the CARE sample. It includes clients in east coast and mid-western US residential and outpatient treatment settings, as well as a detention center and a prison for young, violent offenders. Also included in the study is a group of youth with no or only mild problems. Ages in the youth sample ranged from 2 to 19. Ethnicity was varied. Fifty-four percent (441) of the sample were Caucasian, 36% (300) were African American, 3% (22) were Hispanic, 1% (7) were Asian, 0.2% (1), and 4% (37) were other ethnicities. Eighteen were missing this data. Gender distribution of the sample was 66% (545) male and 32% (265) female. Fifteen were missing this data. Slightly less than half (46%, 375) the sample had histories of assaultive behavior, while 54% (442) had no history of assaults. Eight youth were missing data in this category. Seventy seven percent of the youth lived in their communities and 22% were in placement. There was a fairly normal distribution of IQ scores. Fourteen percent (114) were reported to have committed assaults in the prospective part of the study. Follow-up was 1-6 years while the youth participated in an outpatient program.
The validity of the instrument was assessed. The total CARE score was significantly correlated with the severity of behavior problem history. The mean score of those with no history of assaults was 1, while the mean score of all assaultive persons was 31. The correlation between total CARE score and past assaults was .58 at a significance level of .00 (n = 825, ROC = .84).
Scale reliability was assessed. The Alpha and parallel forms reliability for the scale is .85. Split half reliability ranges from .54 to .85. A regression analysis was applied to the variable Assaults to determine the strongest statistical predictors. Predictors of assaultiveness included Chronic assaultiveness (t = 2.94, p < .01), assault of an authority figure (t = 2.14, p < .05), and moderate to severe behavior problems before the age of 12 (t = 2.00, p < .05).
The population of assaultive persons was subjected to descriptive analysis, also. Sixty-seven percent of the assaultive group had experienced childhood trauma. Seventy-one percent of families of these youth had inappropriate discipline practices. Fifty-nine percent had learning problems, while 70% had school behavior problems. Eighty-three percent had psychiatric problems. The majority had social skill (77%) and anger management (92%) deficits. They often had difficulty relating to pro-social peers, as well.
The CARE showed great promise in helping professionals assess the risk of behavior problems, including aggression. A plan of intervention can also be generated. The total CARE score is significantly correlated with most restrictive placement (r = .33 - .57). This correlation is low because the evaluations in which they were administered sometimes resulted in placements after the evaluation.
References
Seifert, M.K. (Winter 2000). Juvenile violence: An overview of risk factors and programs. Reaching Today�??s Youth, 4:2, 60-71.
Seifert, M.K., Phillips, S., and Parker, S.M. (Fall, 2001). Child and adolescent risk for violence (CARV): A tool to assess juvenile risk. The Journal of Psychiatry and Law, 29, 329-346.
Quinsey, V.L., Harris, G.T., Rice, M.E., & Cormier,