couple dateFlickr/gareth1953
A new study published in The Journal of Sex Research has concluded that the sooner a couple starts having sex, the lower the quality of their relationship.
Perhaps not surprisingly, several media outlets have picked up on this and are publishing headlines along the lines of “First-Date Sex May Harm Couples.”
However, a closer look at the research reveals that both this study and another one that came out earlier this year suffer from the same set of limitations and, in actuality, they really tell us nothing about the effects of early sex on relationships.
The new study, conducted by researchers at the School of Family Life at Brigham Young University, included a sample of over 10,000 individuals currently involved in “serious or steady” relationships who completed an online survey [1].
Participants were asked questions about when they started having sex with their partner and completed several measures of relationship functioning (e.g., satisfaction, communication, etc.). Participants were then lumped into one of four groups based upon timing of first sex: Predating Sex (i.e., hooking-up before becoming a couple; 9.9%), Early Sex (i.e., sex on the first date or two; 35.5%), Delayed Sex (i.e., sex after a few weeks; 47.9%), and No Sex (i.e., couples who were still abstaining; 6.6%).
Results revealed that people who waited longer to have sex scored the highest on all measures of relationship quality. Based upon these findings, the authors concluded that dating couples who have sex therefore have “poorer” outcomes than couples who abstain and that timing of sex represents an important “turning point” in the relationship.
However, my reading of this research suggests that the conclusions drawn are not warranted by the data. First, the results showed that all couples, no matter when they started having sex or how long their relationship had lasted, were satisfied on balance and had positive communication (i.e., average scores for each outcome were above the midpoint of the scale).
Thus, none of the groups had poor quality relationships. All these data tell us is that people who delay sex report being slightly (and I mean very slightly) happier—they do not say that people who have sex early are unhappy. Also, the authors suggest that this minor happiness difference might affect how long these relationships ultimately last, but the study does not provide any evidence of this because it was not longitudinal in nature.  
Second, this is correlational research, meaning that although early sex was related to lower relationship outcomes, we don’t know why. It may have nothing to do with sex at all. For example, perhaps there are personality differences between those who have sex sooner and those who abstain that could explain the observed pattern of effects.
Finally, while the findings were statistically significant, their real-world meaning is highly questionable. Consider that the overall difference in satisfaction and communication scores between those who had early sex and those who abstained was one-tenth of one point on a 5-point scale. Does one-tenth of one point mean that having sex on the first date will necessarily kill your relationship? No. If you can honestly look at these data and come to that conclusion, then please contact me because I have a nice bridge I would like to sell you.
The real reason these findings are statistically significant is because the researchers collected such a large sample that they were practically guaranteed to find significant results. For those of you who don’t have much of a statistical background, it's a basic fact that the bigger the sample you collect, the more likely you are to find statistical differences between groups.

However, if you can only find such a difference with a huge sample (and 10,000+ people is enormous as far as samples goes), it means that the groups actually aren’t that different to begin with. For those of you with more of a statistical background, you’ll understand what I mean when I say that the effect sizes were small—and I mean small. Partial eta-squared (a standardized metric for determining the size of the effect, where .01 indicates a small effect, .06 medium, and .13 large) ranged from .006 to .01 in this study. Thus, we are dealing with miniscule effects that emerged as statistically significant, but just have no real world implications.
To sum it up, this study simply does not provide any evidence that abstaining from sex is a better recipe for success than having sex whenever you and your partner feel most comfortable, so don’t be duped by any headlines to the contrary.
[1] Willoughby, B. J., Carroll, J. S., & Busby, D. M. (in press). Differing relationship outcomes when sex happens before, on, or after first dates. The Journal of Sex Research. doi: 10.1080/00224499.2012.714012