Attrition rates at the 3-year follow-up for the CD-ROM intervention, the CD-ROM plus parent intervention and the control groups were 7.9%, 11.8% and 6.7%, respectively. Analyses of pretest scores of those who did or did not complete the follow-up failed to show differences.
Intervention delivery data from observations, parent reports and monitoring revealed consistently high levels of fidelity in both intervention groups. In the CD-ROM intervention and the CD-ROM plus parent intervention, 95% and 91% of youths, respectively, completed both the initial and booster interventions. As for parents, 163 (83%) watched the videotape, 131 (67%) attended the workshop and 155 (79%) completed the interactive CD-ROM exercise with their adolescent children.
Analyses of pretest data failed to show differences between groups on any outcome variable. Repeated measures MANOVA were subsequently performed. Beginning with family involvement outcomes, these analyses found significant main effects for intervention (F = 34.94, 2/1,320 df, p < .001) and for measurement occasion (time) (F = 418.24, 4/5,280 df, p < .001) and intervention by time interactions (F = 60.79, 8/5,280 df, p < .001). Results of comparisons within time measurements and across measurement occasions within study group for family involvement and all outcomes are reported in .
Youths’ scores on outcome variables1 by group and measurement occasion
On the family involvement variable, youths in CD-ROM plus parent intervention group had higher (better) scores than youths in the control group at posttest and at 2-year follow-up. Both intervention groups scored higher than controls at the 1-year follow-up. By the 3-year follow-up, youths who received CD-ROM plus parent intervention performed better than did those who received the CD-ROM intervention alone, who in turn outscored the control youths. Over time, family involvement scores increased from pretest to final follow-up for both intervention groups, whereas the scores for the control youths were stable.
Comparisons of peer influence outcomes revealed significant main effects for intervention (F = 35.70, 2/1,316 df, p < .001) and time (F = 34.93, 4/5,264 df, p < .001) and intervention by time interactions (F = 21.21, 8/5,264 df, p < .001). Youths in either intervention group scored lower than control youths in their susceptibility to peer influences at posttest and at 1- and 3-year follow-ups. Peer influence data 2 years after posttest showed that the CD-ROM plus parent intervention youths outscored CD-ROM-only youths who outscored control youths. From pretest to 3-year follow-up, peer influence outcomes improved (i.e., scores decreased) in both intervention groups while these outcomes worsened (i.e., scores increased) in the control group.
For alcohol use in the past 30 days, main effects were seen for intervention (F = 25.85, 2/1,316 df, p < .001) and time (F = 80.20, 4/5,264 df, p < .001) and for interactions of the two (F = 14.43, 8/5,264 df, p < .001). Youths in either intervention group reported less monthly alcohol use than control youths at posttest and at 1- and 2-year follow-ups. At the 3-year follow-up, youths who received the CD-ROM plus parent intervention reported less monthly alcohol use than youths who received the CD-ROM intervention only, who themselves reported less use than control youths. Reports of 30-day alcohol use by all groups rose from pre-test to final follow-up, but the rise was less marked for youths who received either intervention than for control youths over the 3-year measurement period. Unlike youths in the intervention groups, those in the control group reported steadily greater use of alcohol at every measurement after pretesting.
Analyses of youths’ self-reported cigarette use during the past 30 days found main effects for intervention (F = 103.37, 2/1,316 df, p < .001) and time (F = 47.96, 4/5,264 df, p < .001) and for intervention by time interactions (F = 9.28, 8/5,264 df, p < .001). Cigarette use was lower among youths receiving either intervention than among control youths at posttest and at 1-, 2- and 3-year follow-ups. Over time, cigarette use reports increased in all three groups, with the greatest rises by control youths.
Reported marijuana use in the past 30 days also differed by intervention (F = 131.88, 2/1,316 df, p < .001) over time (F = 62.18, 4/5,264 df, p < .001) and by interactions of intervention by time (F = 9.34, 8/5,264 df, p < .001). Marijuana use was less for youths who received the CD-ROM plus parent intervention than for control youths at posttest. At 1-, 2- and 3-year follow-up measurements, reported 30-day marijuana use was lower in both intervention groups than in the control group. Although all three groups of youths reported more use of marijuana with time, youths in the control group reported greater marijuana use with each measurement occasion following the posttest.