Surveys designed to measure the effectiveness of virginity pledges may be unreliable because adolescents wrongly report their sexual behavior depending on their social circumstances, according to a study by a Harvard School of Public Health (HSPH) researcher.

According to study author Janet E. Rosenbaum ’98, an HSPH doctoral candidate, 50 percent of adolescents who reported on an initial survey that they had signed a virginity pledge—a promise not to have sex before marriage—denied having signed this pledge on a second survey a year later.

10 percent of those reporting sexual experiences on the first survey denied having had such experiences on the second survey and said they had signed virginity pledges since taking the first survey.

Rosenbaum said these inconsistencies seem to result from changes in the participants’ social circumstances or religious beliefs.

“Changes in social circumstance such as having had sexual experiences or leaving born-again Christianity resulted in respondents recanting on their previously reported virginity pledges,” said Rosenbaum. “People answered based on their current social circumstance.”

Rosenbaum said her study suggests that self-reporting about sensitive subjects such as sexual experience can lead to unreliable results. Changing one’s answers to a sexual experience survey may also have medical implications, Rosenbaum said.

“If there are people who truthfully reported having had sex on the first survey, as I imagine there are, when they lie about their past experiences, there is a risk that they are underestimating their possible sexually transmitted diseases,” said Rosenbaum.

Janice S. Crouse, a Senior Fellow at the Beverly LaHaye Institute, which works with the Concerned Women for America, wrote in an essay criticizing Rosenbaum that her research “purports to show that virginity pledges are ineffective,” thereby furthering the “left’s agenda.”

“Clearly, the left is at it again­­—taking or manufacturing any possible ‘evidence’ in order to slam abstinence programs,” Crouse wrote.

But Rosenbaum said the study should lead to new approaches in assessing the effectiveness of virginity pledges, and not to judgments on the effectiveness of these pledges.

“Evaluations of virginity pledges need to use objective measures,” said Rosenbaum. “STD tests are a possibility.”