One of the perks of being the head of eHarmony Labs is that I get to talk to a lot of people about the research behind eHarmony. It also means that I frequently meet people who met their partner on eHarmony or who have a close friend or relative who did.
It’s always cool to meet eHarmony couples, but we’re now responsible for so many marriages in the United States (nearly 100,000 per year) that it’s no longer surprising when I do meet them.
Several of our larger competitors are also responsible for a lot of marriages in the United States. That’s not terribly surprising to me either. Online dating has become the third most-popular way for newlyweds to have met over the past five years. (Interestingly, for people over age 50, it was the most-popular way to meet and get married. But that’s the subject of another discussion.)
As impressive as this is, my colleagues and I aren’t doing our jobs correctly if all we’re doing is bringing people together. We want to give the couples we match a strong foundation that can help them stay together happily over the long haul. Ultimately, our customers won’t benefit if they’re in a mediocre or bad relationship. The greatest psychological, health and other benefits come from being in a good, satisfying, and long-lasting relationship.
It’s not about creating a lot of relationships; it’s about creating a lot of good relationships.
So lately we’ve been asking ourselves a basic question: Does where you meet your spouse make a difference in how happy the marriage is and if the relationship lasts?
Interestingly, we couldn’t find any studies that investigated whether how married couples meet makes a difference on the outcomes of those relationships. So we conducted two of our own with Opinion Research Corp. to begin to get an answer.
The first study looked at relationship satisfaction. It was an online survey of 7,386 adults who got married in the past five years. (The sample size and selection criteria mirrored a study Match.com conducted last year.) We asked each of these adults how they met their spouse. Then, we gauged their relationship satisfaction using the Couples Satisfaction Index (CSI), which was developed by Janette Funk and Ron Rogge at the University of Rochester and is one of the best relationship satisfaction measures in use today. The higher people score on the CSI, the more satisfied they are with their relationship.
It turns out that eHarmony couples are more satisfied with their marriages than those who met on Match.com, via friends or family, or at a bar/social event. (Match.com was the only other online dating site included in this study since they have made a large number of marriages in each of the past five years.) These differences were statistically significant even after controlling for how long the couples had been married; the respondents’ age, gender, education, and income level; and whether they had children or were previously divorced.
People who met their spouse on eHarmony also reported higher relationship satisfaction than those who met their spouse at work or school, but that difference was only marginally significant. People who met their spouse at church or a place of worship were just as happy as those who met on eHarmony, but keep in mind only about 6% of people meet their spouses that way.
We also asked participants if their relationship had “lost the spark,” since a loss of chemistry can be a precursor of relationship dissatisfaction. Guess what? People who met on eHarmony were least likely to feel their relationship had gone stale compared to all other methods of meeting except church or a place of worship. Again, all the same controls were used – age, length of marriage, etc. (Higher numbers indicate a greater likelihood of losing the spark.)
So far, so good. But now we get to the big question. Does how you met your spouse impact how likely you are to get divorced?
Of course all marriages have their ups and downs. One of the more interesting lines of research over the last two decades has been to figure out why some couples endure and others get divorced. Dr. John Gottman at the University of Washington and Dr. Robert Levenson at UC Berkeley found that discussion of divorce is one of several key indicators of eventual divorce. (Both are among the best marriage researchers in the world.)
So we asked people whether they or their spouse had ever seriously suggested getting separated or divorced.
It turns out that the proportion of couples who discuss divorce doesn’t differ widely across the various ways couples met. Even though eHarmony couples were the least likely to discuss divorce, they were not statistically different from those who met at church/place of worship, work/school, or through family/friends. But there were statistically significant differences between the eHarmony couples and those who met at bars/social events and those who met on Match.com. (We used the same controls in this analysis as we did before).
But this isn’t divorce; it is only a leading indicator of divorce. So in another study we looked at people who actually got divorced and how they had met their partners.
We have wanted to do this study for a while now but it was tricky for a few reasons. Online dating sites have only been around for a little more than a decade, and only really been popular for about the last 7 or 8 years. That isn’t much time for a large number of couples to meet, get married, and then get divorced.
One way to do this study would be to sample from the population at random and hope to find a good sample of people who had gotten divorced to a spouse they met on an online dating site. That means of all the people in the United States, we would be looking for those who got married in the last decade AND who met their spouse on a particular online dating site AND who got divorced from that person. That would have been a lot of phone calls!
We finally found a better way to do this study. We identified an online panel of 4,000 people who had been married to AND divorced from that person in the last 15 years. With Opinion Research Corp. we surveyed this group and asked them where they had met the person they had divorced.
We concentrated on marriages that started in the years 2005 to 2009. We did this because we started to track how many eHarmony marriages occurred each year with Harris interactive in 2005 (with follow up studies in 2007 and 2009). And in those studies we asked people who hadn’t met on eHarmony how they had met their spouse.
We were also able to analyze Match.com couples because they reported how many marriages they were responsible for during this period. Good estimates of the percentages of marriages attributable to each of the most popular ways couples met during those years were available.
Our final sample was 506 people who had gotten married between 2005 and 2009 and subsequently gotten divorced within the same time span. We had hoped for a larger sample, but it is still the biggest study of its kind. (We should note that because this study was conducted online, there were probably more divorces from couples who met using an online dating sites than are actually out there in the population.)
The preliminary findings are interesting.
As it turns out, in most cases, the expected number of divorces was very close to the actual number of divorces we observed in the sample. So it didn’t really matter how you met your spouse, you were just as likely to get divorced. The big differences were people who met on eHarmony (66.6% less likely to get divorced), those who met through school (41.1% less likely to get divorced), those who met through some other means (16% more likely to get divorced), and those who met at a bar (24% more likely to get divorced).
We determined the number of people in the sample who should have met on eHarmony, Match.com, via a friend or family member, etc. by looking at the percentages of marriages that were due to each different method of meeting for each different year. This was based on our own studies with Harris interactive or the study Match.com reported last year. For example, we knew that eHarmony was responsible for 0.74% of all US marriages in 2005. So of all the couples who got married in 2005 and later divorced, eHarmony should be responsible for 0.74% of them. This means the number of divorces we expected to find was based on the percentage of eHarmony marriages in the United States and the year those couples got married. (For statistics wonks, here are the numbers.)
When we added everything up we expected to see nine eHarmony divorcees out of the 506 we had in our sample. There were three. So in this sample, eHarmony couples had a 66.6% lower risk for divorce than would have been expected given eHarmony’s share of marriages in the population. Here’s what it looks like if you analyze the data this way:
We realize the numbers of eHarmony divorces is pretty small and this is only one sample of divorces. We don’t know if these results will replicate in another sample or generalize to all marriages. Those are important limitations to this study that need to be acknowledged. We’re already working on replicating these findings to address these limitations.
Another important thing to remember is that these studies only show us WHAT happened, not WHY it happened. There are lots of potential reasons for these findings. For instance, the lower risk of divorce and higher satisfaction observed for eHarmony couples could be the result of a selection bias among our user base, or it could be attributed to our compatibility matching process, or it could be some other reason entirely. Couples who met at school could be less likely to get divorced because they have similar backgrounds or because they have more time for their personalities and interests to develop together. We’re going to try to dig into those reasons over the next couple of years and we’ll publish those findings.
There is one last important thing to keep in mind. How you met your spouse is only one of many reasons for why a couple eventually ends up unhappy or divorced. Many relationships that start off shaky end up lasting a lifetime. Others that have a great foundation still end up in trouble. How you meet is only the starting point. You, and your spouse, control where you end up.