Study 1
Foshee and colleagues (2005) examined five waves of data collected on a group of students participating in a quasi-experimental evaluation of the Safe Dates program reported earlier by Foshee and a different set of colleagues (1998) that was conducted during 1994 and 1995. The evaluation took place at 14 public schools that had eighth and ninth grades. The schools, which were in a primarily rural North Carolina county, were stratified by grade and matched on size. The schools in each pair were randomly assigned to a treatment or to a control condition.
In the original study, baseline measures were completed by 1,886 youths (80 percent of the 2,344 eligible participants). The sample consisted of 1,566 adolescents who completed the baseline questionnaire and who were in either the control group or the Safe Dates program. Students in the original study who received Safe Dates plus a booster condition at 1 year were excluded from the longitudinal study. Of the 1,566 adolescents, 72 percent were white and 46.8 percent were male. Mean age at baseline was 13.9 years. There were 636 youths in the treatment group and 930 in the control group. The exclusion of the adolescents in the original Safe Dates group who received the booster created the imbalance between the treatment and control groups. Baseline equivalence of the groups was assessed, and no significant differences were found in outcome, mediating, or demographic variables between the treatment and control groups.
The Likert-scale questionnaires used in this evaluation were designed specifically for this study to measure four victimization and four perpetration variables. Psychological abuse perpetration was measured by asking: “During the last year how often have you done the following things to someone you had a date with?” Fourteen acts were listed (e.g., damaged something that belonged to the other person, insulted that person in front of him or her). Response options ranged from 0 (never) to 3 (very often). A parallel set of questions was used to assess psychological abuse victimization.
Eighteen additional questions were used to measure the following violence variables: moderate physical abuse (e.g., scratching, twisting partner’s arm); severe physical abuse (e.g., burning, choking, beating up); and sexual violence. Responses ranged from 0 (never) to 3 (10 or more times). Parallel questions were used to measure moderate physical abuse victimization, severe physical violence victimization and sexual dating violence.
Follow-up data was collected from treatment and controls at 1 month (wave 2), 1 year (wave 3), 2 years (wave 4), 3 years (wave 5), and 4 years (wave 6). The present evaluation reports up to wave 5 (3 years). There was a 50 percent rate of attrition from wave 1 to wave 5 but no group differences in attrition. Data missing because of attrition was handled using multiple imputation procedures. According to this procedure, sets of plausible values for missing observations are created on the basis of a specified missingness equation and an algorithm that preserves uncertainty about nonresponse.
The final data analysis was conducted using random coefficient regression analysis, using a nested error structure, accounting for correlation within individuals’ responses over time and of the individual responses within schools.