School-based programmes that seem to work: Useful research on substance use prevention or suspicious stories of success?
Pape H. Nordic Studies on Alcohol and Drugs: 2009, 26(6), p. 521–535.

According to a commentator, this "trenchant critique" of the evidence for school-based alcohol and drug prevention curricula is "unfortunately, largely on target". The focus is on methodological concerns which might undermine positive findings, and on whether these survive a programme's transplantation to real-world conditions.

Summary
School-based prevention programmes targeted at adolescent substance use rarely seem to have the desired effects on behaviour. Some outcome studies do conclude that such programmes have been successful, but they are relatively few. Nevertheless the body of published research in this field may originate from unrealistic optimism due to publication bias and underreporting of no or counterproductive effects. Anecdotal evidence suggests that the literature is biased in favour of studies with positive findings.

Moreover, nearly all the studies have been carried out by programme developers, and it is well known that researchers with vested interests are more likely to bring 'good news' than independent researchers. Rather than approaching the field with critical reflection, some evaluators have intended to prove that school-based prevention works and have conducted their research accordingly. Examples of questionable analytical approaches and selective reporting of positive findings are consequently not hard to find.

The external validity of evaluation studies with favourable outcomes is also often questionable because almost exclusively they have assessed the effects of programmes delivered under optimal rather than real-life conditions.

In conclusion, the empirical 'evidence' in favour of school-based substance use prevention programmes is generally weak and does not permit a recommendation for the widespread dissemination of any specific programme. School-based programmes might instead be diverted from the unrealistic objective of reducing pupils' substance use to factual teaching. A more appropriate target for education which seeks to reduce alcohol-related harm might be to engender support for truly effective prevention policies such as raising prices and restricting availability.