Who Needs Evidence?
From The Irrelevance of Evidence in the development of
school-based drug prevention policy , 1986-1996D.M. Gorman, in Evaluation Review, vol 22, (1), feb. 1998 118- 146
Kinder, Pape and Walfish (1980) reviewed evaluation studies from the late 1960s and early 1970s, most of which were concerned with information-based programs. As with earlier reviews, they concluded that these programs were ineffective in reducing drug use and might even serve to exacerbate the problem. Goodstadt (1980), addressing the issue of counterproductivity in greater detail, concluded that the available evidence indicated "that 'negative' program effects were not an isolated phenomena, but occur frequently enough and affect self-reported behavior often enough to require more careful scrutiny" (p. 94).
Thus, by 1980 there was little evidence available from program evaluations to support the idea that school- based education was among the "essential components" of a comprehensive drug control strategy. Indeed, in the opinion of many researchers, such education was apt to do more harm than good.
[There are 6 studies that discuss application of social influence programs to illicit drugs]
p. 123 The findings of the two favorable studies are far from compelling.
[details omitted] The other four studies shown in Table I found no statistically significant differences in patterns or levels of illicit drug use between recipients of social influence programs and comparison subjects at follow-up.
The bulk of the available evidence therefore indicated that social influence programs were little better than earlier programs. In short, by 1986, when the federal government committed more than $200 million to school-based programs to fight illicit drug use, evidence indicating effectiveness of this strategy was almost nonexistent.
Differences between ALERT subjects and comparisons were nonexistent at the 10th and 12th grade follow- up. Ellickson, Bell and McGuigan (1993) attribute this to the absence of booster sessions in the schools after the first year of the program, and call for additional research to develop and test such efforts. This ignores the fact that the short-term effects of ALERT were minimal. Why would high and moderate risk subjects benefit from more of the program? Why is more evaluation required? As Ellickson (1995) observes, booster sessions are intended to "extend program effects." For ALERT, there were essentially no program effects to extend.
The latter aspect of the ALERT evaluation illustrates a peculiar feature of school-based drug prevention research during the past 10 years Whatever the outcome, the recommendation is for more of the program and more evaluation. With the exception of D.A.R.E., negative findings are seldom accompanied by a suggestion that we try something else.
Information and affective programs of earlier years were unable to survive negative evaluations; in contrast, social influence programs invariably live to fight another day.
p. 139 The evidence presented herein, from both national surveys and program evaluations, shows that
we have yet to develop successful techniques of school-based drug prevention. The claims made on behalf of this aspect of the nation's drug control policy are largely unsupported by empirical data. Evidence is cited selectively to support the use of certain programs, and there is virtually no systematic testing of interventions developed in line with competing theoretical models of adolescent drug use.
p.141 The question remains as to why policy makers champion drug prevention programs that have so
little grounding in empirical research.
In considering this, it is instructive to recall that for close to 30 years, Soviet agricultural policy was developed in accordance with the theories and research of Trofim Lysenko. According to Lysenko's theory of inherited acquired characteristics, it was possible to transform one crop into another (e.g. wheat into rye) through changing its environment (e.g. by planting it in a different season). Lysenko's "science" thrived under Stalin's regime, in the face of disastrous consequences, as it was totally in accord with the prevailing political philosophyresearch data were irrelevant.
Similarly, the belief that school-based programs can teach children the skills to be "drug free" is entirely in keeping with the individually oriented, zero-tolerance orthodoxy of current U.S. drug control policy. The programs thrive not because research demonstrates their efficacy and superiority over competing approaches, but because the principles upon which they are based are compatible with the prevailing wisdom that exists among policy makers and politicians.
And, judging from recent government publications and the viciousness with which critics are attacked, the uncritical acceptance of school-based social skills training seems likely to continue into the near future.
Comments to firstname.lastname@example.org