Related articles and links |  1 of 2

A new study on "virginity pledges" suggests that they are ineffective and perhaps dangerous. Should we rethink how we approach teenagers about sex?

This month's issue of Pediatrics includes a study by Janet Rosenbaum of Johns Hopkins' Bloomberg School of Public Health on the effects of "virginity pledges" — public vows to abstain from sex until marriage. Using a newer, more effective method to analyze data from the National Longitudinal Study of Adolescent Health (Add Health), Rosenbaum finds the mere act of taking a public vow of abstinence ineffective and likely to lead to riskier choices if (or when) teens have sex.

The study found that teens who took a virginity pledge have sexual relationships that are nearly identical to those of similar teens who did not make such a pledge. The one area in which the pledge does make an impact is negative: Teens who took a virginity pledge and did have sex were less likely to use condoms and other forms of birth control.

The bottom line: Taking a virginity pledge is at best ineffective, and may even be dangerous for the health of those who break their pledge.

The study also has a warning. By focusing on virginity pledges, parents, churches, and sex educators may not only be wasting resources, but may actually be causing harm as well. Those who take virginity pledges — indeed, because they take virginity pledges — are more likely to have unprotected sex when they do have sex. The reason for this unintended consequence is unknown. It is unlikely due to a lack of sex education in school, as those who take a virginity pledge report the same type of health and sex education as other teenagers. Perhaps (and this is only a hunch) when a teen takes a virginity pledge, parents consider it a fait accompli ("mission accomplished"), and thus don't provide continued sex education to their children. What the Rosenbaum study finds, however, is that there may be more that teens need to learn.

When she began the study, Rosenbaum was interested in addressing the following question: Does taking a vow to abstain from sex alter sexual behavior? To understand how Rosenbaum answers this question, it's key to keep in mind how not to answer it. The wrong way for an academic to probe this question is to compare those who take vows with those who do not, but Rosenbaum does not fall into this trap. Those who take abstinence vows do so voluntarily and are very different from the typical American teenager. The Add Health survey reveals they are more likely to be young, female, and from religious families. These teens also value their faith and are active in church. They are more likely to be "born-again" Christians and to have a high view of the Bible. Because of these demographic and religious differences, perhaps teens who take pledges have negative views toward sex and birth control. In short, those who take virginity pledges are the kind of teens who would be more likely to delay having sex until marriage anyway, even if they did not take a pledge.

What makes Rosenbaum's study an important step forward is how she deals with this problem. Previous studies using the Add Health data considered some of the factors that made pledgers different from other teenagers. Yet there were many, many other factors these studies could not take into account due to limits of the statistical models used at the time. Rosenbaum uses a newer, more appropriate method for this type of research called "propensity score matching." The method has become very popular in the social and behavioral sciences because it allows the researcher to have much more confidence that the differences in outcomes (in this case, of sexual behavior) are due to different "treatments" (taking a pledge or not). Rosenbaum is able to test whether virginity pledges per se make a difference in sexual behavior, or if the differences between pledge takers and all other teenagers are due to some other cause. Here is a thumbnail sketch of how she conducted her study.

next page... |  1 of 2

share this page

January (Web-only) 2009, Vol. 53