Social networks and peer education.

Abstract

By testing the social-network method of group assignment and peer leader selection in a randomized trial design with a large number of classrooms and schools, Valente et al.1 made an important contribution to the field of peer-led adolescent tobacco use prevention. My colleagues and I recommended such a research design as a follow-up to our similar 2-year pilot research project2–7 with 347 sixth-grade students in 7 schools in which we tested the effectiveness of using social-network analysis to select peer leaders and to form groups for instruction in a peer-led curriculum to prevent smoking. We compared (1) peer-led education in groups formed by and with peer leaders selected through dendrograms based on a computer-algorithm cluster analysis of students’ nominations on a sociometric questionnaire; (2) classmates taught by model students chosen by school teachers and principals; (3) students taught by adult teachers; and (4) a no-treatment comparison group. The report by Valente and colleagues supports our findings2 suggesting that a curriculum taught by influential adolescent peer leaders within students’ social networks could improve the effectiveness of peer-led school health education to prevent smoking. We also concluded, as Valente and colleagues pointed out, that there was a need to further study the informal diffusion of peer leaders’ influence, gender influences, and the influence of peer leaders who practice the behavior a curriculum is aimed at preventing.2,3,5 In addition to replicating our social-network procedures through the use of a large, group-randomized design, Valente and colleagues might have further extended the field of social-network research and its application to the prevention of adolescent tobacco use had they also built upon our findings and recommendations. Considering the precedent established by our research, first presented 20 years ago6,7 and published 13 years ago,2 not only were Valente and colleagues remiss in failing to cite our study, but their statement that “there have been no studies to evaluate how these leaders should be assigned to groups”1(p1837) was incorrect and an unwarranted claim of primacy.