The effort, as Gonzaga put it to me recently, is “sort of like a Bell Labs or Microsoft for love
Although eHarmony was the first dating site to offer science-based matching, Neil Clark Warren seems like an unlikely pioneer in the field
Indeed, it may well take a generation before we learn whether the psychological, anthropological 
, or sociological model works best. Or maybe an entirely different theory will emerge. But at the very least, these dating sites and the relationships they spawn will help us to determine whether science has a place, and if so, how much of a place, in affairs of the heart.
Meanwhile, until these sites start sending me better dating prospects, I figured I’d take Neil Clark Warren up on his offer to introduce me to the thirty-eight-year-old single board member he thought would be such a good match for me. But when I asked a company spokesman about him, I was told that he had recently begun seeing someone. Did they meet through eHarmony? My potential soul mate declined to answer.
Even though he earned a Ph.D. in clinical psychology from the University of Chicago, in 1967, he never had much of a passion for academic research-or an interest in couples. “I was scared to death of adults,” he told me. “So I did child therapy for a while.” With a master’s degree in divinity from Princeton Theological Seminary, he went on to Fuller Theological Seminary’s Graduate School of Psychology, in southern California, where he taught and practiced humanistic psychology (what he calls “client-centered stuff”) in the vein of his University of Chicago mentor, Carl Rogers. “I hated doing research,” he admitted, before adding with a smile, “In fact, I was called ‘Dr. Warm.’ ”
With the help of a graduate student, Buckwalter reviewed the psychological literature to identify the areas that might be relevant in predicting success in long-term relationships. “Once we identified all those areas, then we put together a questionnaire-just a massively long questionnaire,” he said. “It was probably close to a thousand questions. Because if you don’t ask it, you’re never gonna know. So we had tons of questions on ability, even more on interest. Just every type of personality aspect that was ever measured, we were measuring it all.”
Today, eHarmony’s business isn’t just about using science to match singles online. Calling itself a “relationship-enhancement service,” the company has recently created a venture-capital-funded think tank for relationship and marital research, headed up by Dr. Gian Gonzaga, a scientist from the well-known ily lab at the University of California at Los Angeles. ”
The company sought out Fisher precisely because its market research revealed that although a large segment of singles wanted a scientific approach, they didn’t want it to come at the expense of romantic chemistry. “On most of the other sites, there’s this notion of ‘fitness matching,’ ” Fisher said from her office in New York City. “You e goals, intelligence, good looks, political beliefs. And with the fourth one, you do. What creates that chemistry?”
According to Fisher, each response is correlated with one of the four personality types: Choice A corresponds to Explorer, B to Builder, C to Director, and D to Negotiator.
But you can walk into a room, and every one of those boys might come from the same background, have the same level of intelligence, and so on, and maybe you’ll talk to three but won’t fall in love with any of them
The goal is to incorporate this information into the algorithm to provide better matches, but it can also serve as an accuracy check of the data. Say, for instance, that Jack describes himself as a fashionable dresser, but Jill reports that he showed up for their date in flip-flops, cut-offs, and a do-rag. If the feedback from a number of Jack’s first meetings indicates the same problem, Chemistry will send him an e-mail saying, “Jack, wear a pair of trousers.”
Schwartz’s Duet model consists of a mere forty-eight questions and focuses on eight specific personality characteristics: romantic impulsivity, personal energy, outlook, predictability, flexibility, decision-making style, emotionality, and self-nurturing style. On the first four, she believes, a well-suited couple should be similar; on the last four, however, a couple can thrive on either similarity or difference-provided that both people know themselves well enough to determine which works best.
This, she said, distinguishes PerfectMatch from eHarmony and Chemistry. “In the Chemistry test,” said Schwartz, who is a friend of Helen Fisher’s and a fan of her work, “there was a question about where you’d like to live. And I chose the country. And I would-but the people I tend to prefer are in the city. So they sent me people from Bass Breath, Arizona. And there was no way I could change it! At PerfectMatch, we don’t overdetermine people’s answers that way.”
While the Winkler clients fill out personality profiles similar to the ones found online, the difference, Ahlin believes, is the hour-and-a-half interview. Some of these matchmakers have a psychological background, but others are recruited for different reasons. “We go for people who have a heart, are good listeners, are empathetic, and who just have a feel for matching people for the long term,” Ahlin told me. “On resumes, we look for evidence of good people skills-PR, customer service, nursing. It’s not necessarily about an intellectual understanding. People either get it or they don’t.”
Moreover, in the future, science-based dating sites will evolve in ways that mimic real-world situations. Galen Buckwalter, eHarmony’s research-and-development head, said that rather than relying on self-reports to assess how comfortable a person feels in social situations, his group is developing a model that will use computer simulation to immerse people in scenarios-a bar, a party, an intimate dinner-where variables like gender composition can be altered. “How does this person interact differently as the variables change?” Buckwalter asked. “I don’t think we’ll be relying on self-report twenty years from now. I think not only will data collection advance, but so will our analysis. We’re just at the beginning, really.”
