Study 1
Stormshak and colleagues (2019) conducted a longitudinal study to examine the effectiveness of Young Adult Family Check-Up (YA-FCU) program on reducing young adult risk behavior. In an earlier study conducted by Stormshak and colleagues (2011), families in Oregon were randomly assigned to either receive Family Check-Up (FCU) for Adolescents [treatment group] or school as usual (comparison group) during sixth grade (in middle school). A total of 593 families were recruited, with 386 randomly assigned to the FCU for Adolescents treatment group and 207 randomly assigned to the business-as-usual comparison group. Participants were followed annually into high school. The original study sample was 51 percent male and 49 percent female, with 36 percent identifying as white, 19 percent biracial, 18 percent Hispanic/Latino, and 15 percent Black.
In the 2019 study, young adults and parents from the original sample were recruited 1 year after high school. Of the 386 participants originally assigned to the intervention condition in middle school, 134 were recruited to participate in YA-FCU. A total of 138 of the 207 originally assigned to the comparison condition in middle school was recruited for the current study. These young adults and their families were not offered the program.
Participants were assessed at an average age of 20 (Wave 1, n = 441) and again at age 21.5 (Wave 2, n = 415). Those who participated in the YA-FCU did not statistically significantly differ from those who did not participate at this wave by sex, race, or substance use.
Parents and young adults were asked to complete a survey that included the Young Adult Risk Scale, which was adapted from an earlier instrument developed by colleagues at the Oregon Research Institute (Metzler et al. 1998) and from the Child and Family Center Youth Questionnaire (Child and Family Center 2001). The young adult risk construct was measured with a summative risk score indexing participants from low to high risk, using items selected to assess markers of a problematic transition to young adulthood. Dichotomous items were used to indicate with yes or no whether a risk marker was present. There were 44 marker items covering the domains of vocational risk (for example unemployment, dropped out of school, never held a paid job); alcohol, marijuana, and illicit drug risk (e.g., alcohol and drug use, drug abuse, addictive drug use, drug interference with normal functioning); sexual behavior risk (e.g., unprotected sex, sex while under the influence); and socio?emotional and behavioral risk (e.g., internalizing and externalizing behavior problems). Approximately 1 year later, all young adults were invited to complete follow-up questionnaires.
An intent-to-treat analysis was used to examine change in young adult risk behavior approximately 1 year after receiving the YA-FCU versus change in the comparison group. The analysis conducted was specified as pre?post autoregressive or analysis of covariance prediction models using structural equation modeling. The results were presented in the form of standardized regression estimates (controlling for pretest levels of total risk). The authors conducted a dose-response analysis to determine whether higher levels of YA-FCU would produce greater reductions in risk.
