Newsweek has a column by Sharon Begley about how easy it is to conduct science that turns out to be meaningless due to poor study design. The article was prompted by the recent news that abstinence only programs don’t change the rate of kids having sex. Some of the flaws mentioned in this column:
* Evaluate only the kids who take a virginity pledge (the ones who agree to the pledge are likely to keep it).
* Exclude from the final tally those kids who have sex.
* Follow a program’s effectiveness for a very short time (“But that’s all we have money for.”)
* Ask kids not whether they’ve had sex but whether they “understand” the psychological and health gains of abstinence.
You don’t need an “agenda” to screw up science, you just need a bit of incompetence. I’ll note, though, that there is a great deal of incompetence done by those to pushing the abstinence agenda.