of a Family-Based Substance Abuse Prevention Program
in Rural Communities
Published in The Journal of Primary Prevention, Vol 18, No. 3, 1998
Baseline and one-year follow-up. In the fall of the 1993 school year, a baseline survey was administered to parents and students in the four evaluation schools in the grade targeted for the FIA program The same survey was administered in the fall one year later. FIA participation was voluntary, thus at the beginning of the year it was not known which families would choose to participate in the program. Identification numbers were assigned to students and parents so that it was possible to distinguish between FIA participants and nonparticipants and to link baseline and follow-up surveys. Those families which participated in the program became the treatment group and those that did not participate became the comparison group.
Students completed their surveys in school, and a make-up day was provided for students who were absent. Passive parental consent was obtained with approval of the Wayne State University Human Investigations Committee and the schools. Students placed their surveys directly into an envelope, which was then sealed by one of the students in order to increase their sense of confidentiality. Ninety-six percent of the students completed a baseline survey.
The one-year follow-up response rate was 71% for students in the comparison group and 74% for student program graduates. The majority of the attrition from the baseline to the one-year follow-up was due to students moving out of the school district. The student survey refusal rate was less than 5% at each school. Students who completed a one-year follow-up survey had higher grade point averages (M = 2.92) than those who did not (M = 2.62), F(1,494) = 9.78, p < .001. They also had fewer absences (M = 3.99) than those who did not return a one-year follow-up (M = 5.46), F(1,501) = 8.64, p < .003. There were no significant differences on any of the outcome measures.
Parents were mailed surveys and returned them in prestamped, pre-addressed envelopes which were sent directly to the evaluators. A parent was defined as the mother, father, step-parent or an adult guardian of the student. At least one parent in 54% of the households completed a baseline parent survey, with females (61%) accounting for more surveys than males (39%). The one-year follow-up response rate was 38% for parents in the comparison group and 69% for parent program graduates. Again females (65%) returned more surveys than did males (35%).
Parents who returned a one-year follow-up survey reported higher levels of education (M = 13.08 years) than those who did not (M = l2.7 years), F(1,441) = 3.86 p < .05. Those who returned a survey also had slightly fewer children (M = 2.28) than those who did not (M = 3.16), F(1,441) = 3.88, p < .05. As with students, there were no significant differences on any of the outcome measures.
Students and parents who returned the surveys were given a small incentive. The school principal was allowed to select the incentive which she or he thought would be most appropriate for that school's population. Students' incentives included a free brownie at lunch or a fast food french fries and pop coupon. Parents' incentives included free milk coupons or a lottery for $50.00.
Participants only: Pretest, posttest, and 10-week follow-up. In addition to a baseline and one-year follow-up survey, program participants completed a pretest (the first night of the program), a posttest (six weeks later during the last night of the program), and a 10-week follow-up survey (mailed one month after the program ended which was 10 weeks after the pretest). The pretests and posttests were administered at the program; the 10-week follow-up was mailed to participants with a pre-addressed, prestamped return envelope. Incentives comparable to those described above were given to participants who returned the 10-week survey. These three surveys included the same questions that were on the baseline and one-year follow-up. The purpose of these additional data points was to examine the short-term effects of program participation. These additional measurements were not attempted for comparison group members because of the time burden. Response rates for FIA graduates on the pretest, posttest and 10-week follow-up were 94%, 88% and 73%, respectively.
Program attrition. Seventy-one percent of the participants who attended Session I of the program graduated from the program (a graduate was defined as someone who attended at least four out of the six sessions). Pretest comparisons were made between parent and student program graduates and those who dropped out of the program on 22 measures. The only significant difference was that parent dropouts reported less family activities (M = 2.90) than did parent graduates (M = 3.12), F(1,82) = 5.45, p < .02. Parents and students who started the program but did not complete it are not included in either the participant or nonparticipant group.
Most concepts were measured with multi-item scales that were included for both parents and students at each time point. A few measures were only asked of one age group and one measure was only included for program participants. Details about the instruments are described below. The Cronbach coefficient alphas were of comparable magnitude at the baseline and one-year follow-up, so they are only presented for the baseline measures.
Family cohesion. Family cohesion was measured with the 9-item cohesion subscale from the Family Environment Scale (Moos, 1986). The scale has a true/false response option and all items were averaged into a single family cohesion score. The cohesion scale assessed the "degree of commitment, help and support family members provided one another" (Moos, 1986, p. 2) and a higher score reflected greater family cohesion. Cronbach coefficient alphas for the students and parents on the cohesion scale were alpha = .81 and i>alpha = .80, respectively.
Shared family activities. For parents only, the amount of time spent in family activities was assessed with an 8-item scale (Sebald & Andrews, 1962). A sample item asked "How often do you participate with your child in activities or hobbies?" The scale provided a 4-point response scale with options ranging from never to often. The 8 items were averaged for an overall family activities score (alpha = .88).
School attachment. School attachment was measured with a 10-item scale from the Effective School Battery (Gottfredson, 1984). The Attachment to School subscale uses a 2-point response option and assesses whether respondents "like" or "don't like" the student's school, teachers, principal, counselors and classes. The items were averaged to obtain a global school attachment score for both students and parents (alpha = .77 and alpha = .64, respectively).
Participation in school activities. A school activities scale was developed by the evaluation staff. Students and parents were asked to report in a yes/no format whether or not they were involved in different activities at the child's school (e.g., member of a club or team, attended a PTA meeting). Students reported on three school activities and parents reported on five school activities. An average score was computed for both.
Peer attachment. For students only, perceptions of friends' supportiveness was measured with a 15-item subset of the Inventory of Peer Attachment (Armsden & Greenberg, 1987). The inventory uses a true/false response scale and the items were averaged to obtain an overall peer attachment score (alpha = .88).
Experience with counselors. The evaluation staff developed a 3-item scale which assessed whether or not the student or parent had talked with a psychologist, a social worker or a school counselor. An average score was calculated.
Curriculum knowledge. An FIA curriculum knowledge scale was developed by the evaluation staff to assess the extent to which participants learned the information presented in the program. This measure was only included for parent program participants at the pretest, posttest, and 10-week follow-ups. This was a 6-item multiple choice test; one item was included for each session of the program. An average curriculum knowledge score was computed.
Appropriate attitudes toward alcohol and tobacco use by minors. Rates of ATOD use are very low among sixth and seventh graders (Botvin et al., 1995; Dielman, Shope, Leech, & Butchart, 1989; Johnston, O'Malley, & Bachman, 1993). Thus, if ATOD use is the central outcome measure it is difficult to find significant effects because scores are highly skewed. Therefore, attitudes about adolescent alcohol and tobacco use were assessed for this study rather than actual use. Past research indicates that youths' attitudes about substance use are highly predictive of later use (Hawkins et. al., 1992; Johnston et al., 1993; MacKinnon et al., 1991). Earlier needs assessments had indicated that rates of illicit drug use were very low in these communities; alcohol was the primary drug used by youth. Therefore, the measure focused on attitudes about alcohol and tobacco.
Both the 5-item alcohol and 2-item tobacco attitude scales were created by adapting items from the "Parents" scale in Program Evaluation Handbook: Drug Abuse Education (IOX, 1988). The scales provided a 4-point response option ranging from definitely yes to definitely no. Students' questions were phrased in terms of their friends (e.g., "Would you be upset if your friend took you to a party where alcohol was being used?"). Parents answered parallel items about their child's use of alcohol and tobacco (e.g., "Would you be upset if your teenager got drunk on a special occasion like a graduation party or New Year's Eve?"). The Cronbach coefficient alphas were .81 and .78 for students' and parents' alcohol attitudes; they were .83 and .72 for students' and parents' tobacco attitudes.
The legal drinking age in Michigan is 21 years of age. Students and parents were also asked at "What age do you think that it is O.K. to drink more than a sip of alcohol?"
Demographic and school information. Students' grade point average and number of school absences were collected from the school. Students reported their age on the survey. Parents' surveys included questions about their education, number of children, and annual household income.