Study
Werch and colleagues (1998) recruited students to their study from the sixth grade at a middle school in a disadvantaged Jacksonville, Florida inner city school district in the fall of 1995. After returning their parental consent form, the pilot study recruited 85 percent of eligible students. Eight additional students did not complete the baseline (either withdrawing from the study, being transferred, suspended, or expelled). The 211 participating students were 85 percent African American, 12 percent white, and 3 percent other ethnicities, with half of the students being male. The mean age of participants at baseline was 12.08 years, with 78 percent of students participating in the free lunch program—an indication of social disadvantage. Additionally, 32 percent of participants reported an immediate family member with a drug or alcohol problem and 65 percent reported not having received any drug or alcohol education in the preceding year. They were randomly assigned by computer to receive either STARS (Start Taking Alcohol Risks Seriously) for Families or to a control condition that received a 15-page alcohol education booklet. For the intervention group, the treatment delivered depended on the risk factors identified during the healthcare consultation, with students receiving between two and nine family lessons. The average number of lessons received by students was 5.5.
At a 1-month posttest, 187 subjects participated, with 147 participating at the 1-year follow-up. Significant differences existed between dropouts and participants in terms of baseline measures of free lunches, age, and alcohol use. At the posttest, the authors tested 88 intervention and 99 control students. At the 1-year follow-up, 73 students in the treatment group were compared to 70 students in the control group using t-tests, chi squares, and analyses of covariance (ANCOVA).
The authors used the 77 item Youth Alcohol and Drug Survey to measure alcohol and drug consumption, in terms of frequency and intensity, as well as risk and protective factors. Additional demographic data were also collected, including age, ethnicity, participation in free lunch programs, exposure to alcohol and drug problems, and other education programs, and parental education level. Subgroup analyses were conducted to determine the potential impact of the program on current drinkers.
