Does waiting to have sex until after getting married really improve the quality of sex? I dismissed the claim when I saw it in an article in Journal of Family Psychology. Hmm, I thought, well at least it will be the best sex a person had ever had.
Besides, who still waits until then? Recent studies suggest that 70-90% of adolescents now have sex by age 18 and most dating couples have sex within the first month of dating. Moreover, even adolescents who pledge themselves to virginity do not differ from closely matched peers in whether they have had sex five years later, although by then, most who pledged virginity deny that they ever made such a pledge.
But even if my skepticism remained, my interest was rekindled when the claims of the study were picked up by the news media and blogs. Aol.com Health declared "Delaying Sex Credited for Making Marriages Stronger." Across the Atlantic, the Economist echoed "Chastity before marriage may have its uses after all."
Professor Dean Busby of Brigham Young University had studied 2,035 married people ranging in age from 19 to 71, married from less than six months to more than 20 years. Dr. Busby claimed that not only did the couples who waited until after marriage to have sex rate their sex life better, they reported that their marriages were more stable and satisfying, and their communication was better than couples who had not waited.
At a Mormon website Busby was even quoted as saying "When couples become sexual early in the relationship ... other areas do not develop as well...Couples who wait spend time talking and sharing their life, getting to know each other in other ways, strengthening their relations and creating greater relationship satisfaction." The website referred to Busby as "Brother Busby," with "brother" being a term used by Mormons to refer to males sharing their faith. Come on, Brother Busby, are you objectively interpreting your data here or merely preaching to us?
I wondered whether Busby had drawn his sample from his own university. Brigham Young University recently made the news when it suspended a star basketball player for the season because he had sex with his girlfriend. The university is owned by the Church of the Latter Day Saints, (the Mormons). Not all of its students are Mormons, but applicants for admission must have a letter from their clergy asserting that they lead a "chaste and virtuous life, including avoidance of pornography, abstinence from sexual relations outside of marriage, and abstinence from homosexual conduct" Once admitted, students must, under penalty of suspension, adhere to an honor code that includes continuing with these prohibitions, as well as abstaining from alcohol, coffee, and tea.
Brigham Young faculty do not escape from the honor code either. They must obtain yearly certifications from clergy that they are "temple-worthy" and they can be dismissed for violating or contradicting church policy.
Yikes, but before jumping to the conclusion that this study was biased either by its sample or by pressures on Busby to uphold the church policy in his interpretation of his data, I sent him a polite note inquiring where and how he recruited his sample. I asked also what other variables correlated with sexual satisfaction after marriage, besides whether the couple had waited until then to have sex. This information was missing in his paper, although it is crucial to deciding about alternative explanations of results. It is customarily provided in a paper so that readers can come to their own conclusions. Unfortunately Busby did not respond and neither he nor his coauthors responded to three subsequent emails. So, we are left with what was reported in the article.
Just who were the people participating in this study? The article states all had completed a 300 question on-line survey and there is no indication that they were paid to do so. Thirteen percent of the participants found the survey by searching for it on the web, 29 percent of were referred to the online site by their instructor in a class, 25% were directed to the site by a relationship educator or therapist, 8% were sent to the site by clergy, 18% were referred to the site by a friend or family member, and 7% were referred by an ad they saw online or in a print. Only 17% of the sample identified themselves as Mormons, but only 17% of the sample identified themselves as not having a religious affiliation.
Recognizing the potential for bias, Busby introduced statistical controls for religious affiliation and confidently declared "we have been able to demonstrate that sexual timing has a unique effect beyond religious involvement." Not so fast. This is a truly unrepresentative sample, both in sampling strategy and its make up, and statistical controls cannot compensate for this.
Just how large were the effects that were claimed, anyway? After controlling for religiosity, length of relationship, number of sexual partners, and education, whether participants had abstained from premarital sex only accounted for less than two percent of the variance in sexual satisfaction after getting married. This is statistically significant, but hardly worthy of the media and blog attention it got. And we were not provided the basic information to see if other variables might be related to both waiting until marriage to have sex and to being satisfied with sex after marriage. Taking this information might make the already small effect disappear entirely.
We might imagine an experiment in which adolescents were randomly assigned to having sexual intercourse before or after marriage, but it would be difficult to imagine teens agreeing to such a study or sticking with the group to which they were assigned, particularly given the difficulty observed with teenagers pledging themselves to preserving their virginity. So, we can rule out a randomized controlled experiment. We are left with survey-or better, longitudinal studies--to attempt to answer the question whether waiting until marriage to have sex improves the sex or quality of the marriage. But in trying to answer this question with these methods, a host of background factors come into play and we are then left to relying on statistical controls in any attempts to sort them.
Statistical controls are imperfect tools for answering causal questions, only as good as our limited ability to identify the relevant variables, to put them in a causal sequence, and to measure them adequately. And statistical controls are unlikely to compensate for a poor sampling strategy that does not yield a sample representative of the population to which we want to generalize.
In the end, I did not learn anything newsworthy about abstinence or sexual timing and subsequent sexual satisfaction from Busby's study. And yes, when authors are insufficiently disclosing of their results, it is certainly justifiable to rely on whatever you know about them to evaluate their credibility.
We are reminded again that we need to be skeptical of what we read in psychology journals. Just how did Busby's failure to report correlations between other variables and abstinence get past the editor? Didn't the reviewers of the paper detect a bias in the strong interpretations being made of weak data? Maybe not. We should cultivate tools for exercising skepticism about what we read because we cannot rely on the journal editors or reviewers or the media and bloggers to be skeptical for us.