better (e.g., Durlak, 1995; Gottfredson, 1997). Programs that provide booster sessions after the initial activity produce more lasting effects than those that do not. These conclusions are based largely on a handful of studies that have compared effects of an instructional program with and without a booster. Botvin et al. (1984b), for example, show that the effects of Life Skills Training (the peer-led version) on self-reports of marijuana use in the past month taken 16 months after the initial pretest are not statistically significantly different from zero for those students in the condition without the booster, but when additional lessons are provided in the following school year to reinforce the initial lessons, effects at 16 months after the pretest are statistically significant and more than doubled in magnitude. However, others have demonstrated both short-term (Eggert et al., 1994) and long-term (e.g., Lochman, 1992) positive effects of one-shot interventions of approximately the same duration as Botvin’s initial program. Clearly, booster sessions are not a necessary ingredient of successful prevention, but their timing appears to be important.

For example, research may discover that the total dosage of prevention messages can be traded off against the timing of the messages. Brief messages delivered closer in time to the situation in which an opportunity to use drugs is likely to arise, or small doses delivered continually over the life span, may be more effective than long messages delivered within a short time frame, as is most often the case in drug prevention classes as they are offered today. Mass media campaigns—television and radio advertisements, billboards, and posters—offer this potential advantage over classroom-based messages. More research is needed to sort out these potential trade-offs between timing and dose.

Reviews have also concluded that the role of the deliverer is important. Hansen (1992) suggested that the training and background of the leader and the fidelity of presentation may be more important than the content of the message. Tobler compared programs delivered by different types of leaders: mental health professionals and counselors produced the largest effects, followed by peers; teachers produced the smallest effects. Tobler (1992:20–21) concluded that the leader must be someone who is “competent in group process, who can enhance the interactional process and simultaneously focus and direct the group. Successful leaders have the ability to act as guides, as opposed to being dominant. They are able to tolerate ambivalence, and know when to remain silent to facilitate true dialogue. They are able to empower adolescents to make conscientious decisions and to encourage freedom of choice and individual self determination.”

Undoubtedly, the content of the message and the characteristics of the leader interact to produce more or less effective programs. Perhaps the provision of accurate information about the consequences of use by a



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement