Online-dating sites have changed romance for millions of Americans. But claims that such dating leads to hordes of newly wedded couples may be fairy tales. EHarmony claims in television or online ads in the U.K., U.S. and Australia that 2% of Americans who got married last year met through its site. But the stat is based on an online survey. Similarly, a Match.com media kit claimed that 12 marriages a day trace their roots to the site, but the company now says it’s inaccurate.
And Markus Frind, chief executive and founder of Plenty of Fish, doesn’t advertise about marriages, but says his site brings about 100,000 marriages a year, a figure based in part on "some study I found online." Companies’ own numbers have gained notice in part because other marriage researchers don’t have a strong handle on the quantitative impact of marriage dating on mating. It’s a surprising void after a decade in which electronic winks and flirts may have proven more fruitful than bar pickup lines. Not all dating sites are offering up numbers of the betrothed. "Self-reported data — even the most joyous kind — cannot and should not be represented as statistically valid," says Herb Vest, chief executive and founder of dating site True.com. The closest thing to an election-quality poll didn’t directly measure online-originating marriages.
The Pew Internet & American Life Project, a research group, surveyed 3,215 adults in 2005, and based on its findings, reported that three million Americans had entered into long-term relationships or marriages with free personals they met on dating Web sites. But Pew didn’t specify how many of those were marriages. And it isn’t planning any major follow-up, in part because Web-traffic monitors report a stabilization or decline in traffic for many major dating sites. One obstacle to further research is that newlyweds are hard for researchers to find. So online-dating companies looking to tout their success as matchmaking yentas have sought novel ways to enumerate marriages.
Match.com and dating services sites has taken the most varied approach. For years, the company counted member reports of marriages — a total of 1,100 by 2002. Then Match shifted gears, claiming in press releases that it "is credited with more marriages than any other site," citing a 2005 poll by Weddingchannel.com. That site has since been acquired by The Knot, whose spokeswoman couldn’t say how that survey was conducted. Match continued citing that survey through at least 2007. Earlier this year, Match.com took a different tack. Parent company IAC, in a fact sheet prepared for advertisers, claimed that a dozen couples got married or engaged off the site each day. Online Dating Insider blog’s David Evans used other numbers from the fact sheet on dates among Match.com users each year to calculate that it would take an average of 1,369 Match.com dates to get married. IAC subsequently pulled that stat from the fact sheet. Match.com spokesman John Walls says the site is interested in calculating an accurate number.
"Being able to say how many relationships you’re responsible for is very important to your brand," he says. Mr. Frind, of Plenty of Fish.com, scoffed at the dozen Match.com for matchmaking nuptials per day, noting that eHarmony claims 10 times that number. He says that his site creates 800,000 relationships each year, according to exit interviews with departing members. He says that works out to about 100,000 marriages per year, based on a study, the details of which he couldn’t recall, that says 10% to 15% of relationships lead to marriage. The only statistic Mr. Frind knows with certainty, he says, is the number of members who have self-reported success stories on his site — now around 2,000. "I don’t want to pay $200,000 to a research company to find out how many marriages I have per year." EHarmony has twice commissioned studies from Harris Interactive to estimate the dating site’s impact on marriages. The first, in 2005, found that 90 people a day married because of the site. The second, conducted in 2007, found that the number had surged to 236. Using Centers for Disease Control and Prevention data on marriages, eHarmony estimated that it could claim credit for 2% of marriages in the country. When first confronting this measurement, eHarmony faced some daunting numbers. Since so few Americans — fewer than 2% — get married in any given year, calling households randomly to find enough to answer such a survey would be prohibitively expensive. Gian Gonzaga, senior research scientist at eHarmony, estimates that it would have taken 1.75 million phone calls to get a big enough sample. So instead, Harris used a massive online panel to recruit respondents much more efficiently. But such surveys introduce potential bias, George Terhanian, Harris’s president for global solutions, says. Those answering questions about online dating and marriages may be more likely to spend time online and share personal details there, possible prerequisites for turning online courtship into matrimony. Harris takes extensive steps to attempt to correct for these biases. Each respondent is assigned what’s called a propensity score, which attempts to measure how much they are over- or under-represented in online panels. The responses of, say, heavy online users would carry less weight than average. Dr. Gonzaga says eHarmony was also conservative in its calculation, for instance by assuming that no one over 54 married because of the site. And he says the site received reports of 5,000 marriages among confirmed members during the year covered by the survey, which means if just one in eight married members shared their marital bliss with their matchmaker, the eHarmony number checks out. For these reasons, Dr. Gonzaga says he is confident in the Harris survey, though he adds that at first "I was doubtful myself." The site is planning to work with Harris to update the number later this year. Academic researchers remain skeptical. "Such claims should be considered as advertising promotions; no more, no less," says Jeffrey Lohr, a psychology researcher at the University of Arkansas.
Source: online.wsj.com
