It happened again on Daily Kos today, in a post about the funding for abstinence-only education that some Senators from both parties put into the health care bill. I have no clue about what’s in the bill and I’m not debating that. Instead, I’m wondering why people continue citing the study published by Dr. Janet Rosenbaum in the journal Pediatrics as the final proof on the issue. It’s happened everywhere from Kos to this message board, where a thread was devoted to it about a year ago.
Dr. Rosenbaum’s study supposedly proved that teenagers who take a virginity pledge are just as likely as those who don’t to engage in risky sexual behavior. The problem is that the study actually says the exact opposite of what everyone claims it says. It says that teenagers who take a virginity pledge are a lot less likely to engage in risky sexual behavior than teens on average. (Unfortunately it costs 25 bucks to read the article unless you belong to an institution that subscribes.)
The study functions by a matched pair design, a statistics trick whereby you match up individuals from two different sets who are most similar–in this case, individual teens who took a virginity pledge were matched up with similar teens who didn’t take the pledge. Because those who took the pledge were sexually conservative, they were matched up with sexually conservative teens who didn’t take the pledge. And it was found that sexually conservative teens had the same rate of nearly all kinds of behavior, regardless of whether they took a virginity pledge or not. However, when the two groups of sexually conservative teenagers were compared to the entire population of teenagers, it was found that the sexually conservative teens had less risky behavior by all measurements. Lest this get too abstract, the specific trais in question were:
So the Rosenbaum study doesn’t say what most of its fans want it say, and to her credit, Rosenbaum herself never claimed to have shown that abstinence-only education didn’t work. That interpretation was grafted on by others, apparently hoping that nobody would bother looking at the actual source of the study. In any case, the crusade against abstinence-only education should probably look for a source that actually supports their claims, since this study obviously doesn’t.