That is evidenced by the fact that prevention messages produced affirmative responses to questions about curtailment and curtailment messages produced affirmative responses to questions about prevention. (4) Perception of realism is likely a critical element of any anti-drug messaging campaign expected to appeal to a teenage audience.

Exactly what do you recommend for the de-marketing communications campaign?

There would seem to be little basis to presume that either a prevention strategy or a curtailment strategy would be more effective than the other. It is most likely that both approaches would be advisable because teenagers who already use drugs are much less likely to respond as hoped to the curtailment messages whereas teenagers who do not already use drugs would be more likely to respond as hoped to the prevention messages. More specifically, teenagers who already use drugs are beyond the point where a prevention messaging strategy would be productive.

On the other hand, it is conceivable that teenagers who do not already use drugs might respond positively to both prevention and curtailment messages since the risks presented as a consequence of drug use would, presumably, be meaningful and effective in that respect. Nevertheless, these questions should have been investigated by the chosen methodology instead of ignored by the chosen design. In principle, the study, as designed, provided little potential insight into whether a prevention or curtailment messaging strategy would be more useful or which respective messages in each category might be the most effective. If the study were to be redesigned, the following might be a better methodology:

Then, the trial should have been repeated successively for al of the messages, always pairing the prevention theme against the curtailment theme. Next, all of the subjects who preferred the prevention messages should have been retested, this time, comparing their responses to all of the prevention messages to identify the most likely prevention message to be effective. Then, all of the subjects who preferred the curtailment messages should have been retested, this time, comparing their responses to all of the curtailment messages to identify the most likely curtailment message to be effective.
Obviously, if there is any evidence that a given teenager population already uses drugs, the curtailment approach would be more useful than the prevention messaging approach. Because the dangers of drug use is a message more relevant to both groups, it might be advisable to rely on that approach wherever one must be chosen to the exclusion of the other. Unless, there is independent evidence to suggest that the population in whom the messages are expected to be used already use drugs or do not already use drugs, it is likely that there is absolutely no basis for limiting the ultimate initiative to either message. Ultimately, the project should have been better or more particularly designed to identify the best of all possible prevention messages and the best of all possible curtailment messages for an anti-drug initiative that targeted both groups with one message designed to appeal maximally to each.