How not to criticise sex ed programmes

Ross Douthat links to a Jennifer Roeback Morse post from nearly two years ago claiming that a handful of studies show that sex education programmes -- and we're not talking about the abstinence-only kind -- don't work.

Except that Morse apparently doesn't understand what a control group is, and Douthat apparently didn't have his critical thinking hat on when he read her post. (Or he doesn't understand what a control group is, either.) Both of the primary studies Morse cites are from the 15 June, 2002 issue of the British medical journal. The first is a meta-analysis that aggregates data from about two dozen randomised controlled trial studies of sex ed programmes. It's actually quite methodologically sloppy, because of the wide range of differences between intervention and control groups. The control groups, for example, range from no sex ed at all of any kind to `standard knowledge based prevention curriculum' or `usual sex education programmes'. The other looks at a specific new programme, and concludes that `[c]ompared with conventional sex education this specially designed intervention did not reduce sexual risk taking in adolescents' (my emphasis). Furthermore, in the brief summary at the end, they add that `[i]mprovements in teacher delivered whole class sex education have some beneficial effect on the quality of young people's sexual relationships but do not influence sexual behaviour'. So this is actually a highly qualified success, not a failure, of the programme in question.

In short, in both studies, both control and intervention groups included students taking conventional sex ed. They are therefore irrelevant to the question of whether or not these programmes work better than no sex ed or abstinence-only sex ed.