Impact Data - Friendly PEERsuasion
Date
Methodologies
4 sites were chosen as demonstration sites; each site was asked to recruit at least 100 girls between the ages of 11 and 14 to participate in the Friendly PEERsuasion programme. To avoid ethical problems of withholding programming, the evaluation employed a delayed entry design. Approximately 40% of girls completing the pre-programme survey were assigned randomly to autumn participation (treatment group) while the remainder were assigned to the control group, which participated in the programme the following spring. The initial sample included 152 girls (treatment group) and 202 (control group). 4 questionnaires were administered to all study participants: the pre-programme questionnaire (September 1988) when neither group had participated, the first post-programme questionnaire (November 1988) when the treatment group had received the Friendly PEERsuasion training, the second post-programme questionnaire (February 1989) when the treatment group had both received the training and had also taught the PEERsuade-MEs, and the third post-programme questionnaire (May 1989), when both treatment and control groups had participated.
Practices
Participation in Friendly PEERsuasion was significantly related (p=0.10) to delayed onset of drinking among participants who reported never having drunk alcohol prior to the programme. Those who had already consumed alcohol reported lower incidence of drinking at the post-programme periods, although this difference was not statistically significant (p=0.12).
The effectiveness of this programme on delaying alcohol use persisted over the study period. Among girls who reported never having drunk alcohol at the pre-programme questionnaire, 36% of the control group reported first use of alcohol on any post-programme questionnaire. The estimated effect of programme participation was a 14 percentage point reduction in the likelihood of drinking during the study period (p=0.02).
The estimated effect of the programme if both groups had participated in the programme during the autumn 1988 would be to halve the incidence of drinking from the actual rate of over 10% to under 5% (p=0.05). Younger girls who participated earlier were less likely to begin using harmful substances during the study period (p=0.06). Older girls reported similar behaviours regardless of earlier or later participation.
Participation in the programme led some girls to report leaving gatherings where others were drinking alcohol (p=0.05).
Evaluators found no evidence of differential attrition between the treatment and control groups. Attrition in both groups, however, was related to 2 background factors: seeing someone selling drugs near the respondent's home and past cigarette smoking. One implication of this fact is that students who smoked cigarettes were less likely to be included in the post-programme survey.
The effectiveness of this programme on delaying alcohol use persisted over the study period. Among girls who reported never having drunk alcohol at the pre-programme questionnaire, 36% of the control group reported first use of alcohol on any post-programme questionnaire. The estimated effect of programme participation was a 14 percentage point reduction in the likelihood of drinking during the study period (p=0.02).
The estimated effect of the programme if both groups had participated in the programme during the autumn 1988 would be to halve the incidence of drinking from the actual rate of over 10% to under 5% (p=0.05). Younger girls who participated earlier were less likely to begin using harmful substances during the study period (p=0.06). Older girls reported similar behaviours regardless of earlier or later participation.
Participation in the programme led some girls to report leaving gatherings where others were drinking alcohol (p=0.05).
Evaluators found no evidence of differential attrition between the treatment and control groups. Attrition in both groups, however, was related to 2 background factors: seeing someone selling drugs near the respondent's home and past cigarette smoking. One implication of this fact is that students who smoked cigarettes were less likely to be included in the post-programme survey.
Attitudes
A lower percentage of autumn participants reported favorable attitudes toward drinking alcohol after completing PEERsuader training than did their peers who had not yet begun PEERsuader training, although this difference was not statistically significant (p=.20).
Source
Letter sent from Sarah Riester to The Communication Initiative on January 7 2004; and "A Profile of the Evaluation of the Girls Inc. Friendly PEERsuasion Program" on the Out-of-School Time page on the Harvard Family Research Project site.
- Log in to post comments