Modern Jackass: The Sex Edition

There's a little blurb in the Roanoke Times which reads:

Teens who lose their virginity earlier than their peers are more likely to steal, destroy property, shoplift or sell drugs than their virgin counterparts, according to one of the first studies to look at what happens in the lives of teens in the years after they start having sex.

The study, reported in last month's Journal of Youth and Adolescence, found that those who had sex early were 26 percent more likely to be in trouble than those who waited, even years after their sexual debut and well into early adulthood.

The average age of sexual initiation varies widely -- between just older than 11 and 17 depending on the school, with an average age of about 15, according to federal data from the National Longitudinal Study of Adolescent Health. Researchers from Ohio State University looked at national survey data for about 7,300 teens from grades 7 to 12, and compared the students' sexual behavior with that of teens within their own schools.

Those who had sex later had delinquency scores 20 percent lower than their "on-time" peers. Waiting had a protective effect.

Several things about this:

1. It doesn't actually name the study or give a source for it. It does reference the data which the study used, but that data is a large data poll which many researchers use, and is sometimes used to exaggerate the importance of a "study" which is, in fact, simply an analysis of data someone else has done.
2. It doesn't indicate whether or not the research is peer reviewed. It turns out that the Journal of Youth and Adolescence doesn't indicate whether or not the research is peer reviewed either. In other words, we don't know if this is peer-reviewed research or not.

I bring this up because it's a common practice for people to use non-peer reviewed "research" and present it as fact, without any real support behind it. In this case, we're dealing with a newspaper reporting on an alleged science story without any peer-review or clear science behind it. In this case, having managed to track down the original piece (abstract can be found here), there's no indication whatsoever that, as stated in the Roanoke Times article, "waiting had a protective effect."

This, in fact, belies a common misunderstanding about how correlational research works: even if you take the research on face value, we don't know if sexual activity causes delinquency, if a tendency towards delinquency causes teen sex, or some other thing entirely. Correlations do not equate to causation. Even more important: there are two crucial flaws with this investigation:

1. While they separated out people who reported having "forced" sex, they didn't separate out those who were unable to consent to it-- they just referred to "intercourse." It's not uncommon for children who have sex with adults to fail to classify it as forced in self-reporting, even though they're not able to rationally consent.
2. dealing with self-reporting is always a bit tricky-- in a society which discourages teen sex, children who are sexually active may be less likely to self-report. Similarly, children who have engaged in delinquent behavior may be less likely to self-report as well. It's very possible that those children who are more willing to report on sexual activity are also more willing to report on delinquent behavior, just as those who are unwilling to admit to one are unwilling to admit to the other.

This long rant is simply my way of saying: don't believe what you read without doing a little investigation-- when someone presents research to you as fact, be skeptical. This especially applies to any claims about what sexual activity may or may not cause.