The Art of Choosing a Romantic Partner (Part One)

New research explores why we tend to choose the same partners over time.

Posted Jun 14, 2017

Source: wavebreakmedia/Shutterstock

Why do we choose the romantic partners that we do? What shapes the choices we make? While researchers have been weighing these question for decades — and philosophers for centuries more — no one has managed to come up with a clear answer yet.  

According to evolutionary psychologists, humans are primarily attracted to sex partners with specific characteristics that will boost reproductive success. This means that females seek potential partners who bear primary and secondary sex characteristics they regard as signs of genetic fitness — deep voices, masculine characteristics, indicators that they are good providers, etc. — while men seek to make themselves as attractive as possible through competition with other males.   

But that's just one perspective on human mating behavior. According to the concept of assortative mating, we are most likely to be attracted to partners with whom we share basic similarities. These similarities can take the form of genetic characteristics — e.g., research showing that men prefer females whose faces are similar to their own — or involve similarities in cultural, ethnic, or socioeconomic factors.

If you're not persuaded by either evolutionary or assortative theories on mating, there is the self-deception model of romantic behavior. According to this model, lasting relationships are formed when people develop positive illusions that make them less likely to recognize the flaws in their partners. As in William Blake's poem, "Love to faults is always blind / Always is to joy inclin'd / Lawless, wing'd, and unconfin'd / And breaks all chains from every mind," cultivating positive illusions can help individuals turn a blind eye to their partners' flaws. In other words, looking at potential partners through rose-colored glasses can help ensure domestic bliss.  

But most people won't just settle on one partner. We tend to acquire multiple romantic partners over the course of a lifetime. Which then brings us to the question of how consistent we are in our mating choices. Do the kind of partners we select as teenagers match up with the partners we might choose when we are older and, presumably, wiser?   

Given the conflicting variety of theories about human mating, the obvious question arises over whether it's even possible to make sense of what is ultimately a highly personal, and often irrational, decision. It's hardly a cliche for many parents to be confused about the romantic choices their children make, not to mention other friends or family members. Even as we grow as individuals, we may find ourselves making different choices about who we might choose to be with. But is it really possible for researchers to understand matters of the heart?   

A new research article published in the Journal of Personality and Social Psychology attempts to answer this question by examining how consistent people really are in the romantic partners they choose. Paul Eastwick of the University of California Davis and a team of co-researchers conducted a series of studies focusing on the different factors that go into romantic pairings, particularly in terms of those stable factors that influence the romantic choices we make — the long-term factors that can cause us to choose similar partners over time.   

According to Eastwick and his colleagues, the stable factors that seem to dominate mate selection include:

     Active stable factors. These are the qualities that individuals use to evaluate potential partners, and by which they in turn are evaluated. Whether it involves physical attractiveness; appealing personality traits such as humor or compassion, intelligence; or evidence of being a good provider, these different factors can help people judge whether a potential partner has high or low value as a mate. Many people may not even be consciously aware that they tend to choose in specific ways, but similarity-attraction effects often lead to "clustering" among current and former partners due to how much they share specific characteristics. Another stable factor that often plays an important role is relationship aptitude. This deals with how effective people can be in forming and maintaining new relationships, depending on their prior relationship experience and personality.  

     Passive stable factors. These stem from the social environment in which we all live, which can also shape mating choices. Overall, we are most likely to meet and date people whom we might encounter on a frequent basis. This includes dating someone who lives in the same neighborhood, goes to the same school, works in the same building (or nearby), etc. Also known as propinquity, it can be a powerful influence in the relationships we choose. People are also more likely to become involved with those who are similar to them in some way — belonging to the same ethnic group or religious congregation, having the same financial background, etc.  

In their research, Eastwick et al. conducted three empirical studies and one simulation study looking at the similarities that occur among individuals' current and former partners, and how active and passive factors can influence mating choices. The simulation study was used to confirm that people selecting mates based on specific qualities would tend to show clustering among current and former partners on those particular qualities even when the number of mates being selected is very small.

But how well does this apply in real-life relationships?

With the first empirical study, 136 undergraduate students provided Internet links to photographs of two or more current or ex-partners, which were then analyzed for similarities that were visible in the images. Participants were told they were completing a survey on how people depict themselves on social media. They were then asked to provide links to Facebook profile pictures of people with whom they were currently involved sexually or had been involved with in the past. Participants also indicated for each partner whether he or she was (a) a “current boyfriend/girlfriend” (N = 45), (b) an “ex/former boyfriend/girlfriend” (N = 128), (c) a “current non-committed partner (e.g., hookup, friend with benefits)” (N = 24), or (d) a “past/former non-committed partner (e.g., hookup, friend with benefits)” (N = 231). On average, each participant provided links to 4.4 partners and provided more than 428 photographs for the study.  

All of the photographs were then rated on three dimensions — attractiveness, masculinity/femininity, and dominance — and were then compared to photographs of the participants themselves. Results showed significant clustering along these traits for both past and present romantic partners as well in comparison to each of the participants. In other words, people tend to choose partners who are similar to themselves in many ways and also tend to be consistent in terms of the type of partner they prefer.

Since the participants and the partners for whom they provided photographs tended to fall within a very narrow age range, Eastwick and his colleagues conducted a second study to see if the results would hold up in a much broader group. Using data from Add Health, a long-term study of adolescent health, 303 men and 271 women were recruited, and each then nominated two or more romantic partners for whom self-report data was available. A total of 1,110 different partners were examined over the course of the study. (Same-sex relationships were excluded for the purpose of the study.)  

Among the Add Health variables that were examined in the study were educational aspirations, depression, intelligence, self-esteem, vitality, delinquent behavior, and religiosity. Demographic factors such as parental income, parental education, and partner racial background were also included. The results showed evidence for partner clustering among the different variables studied, though the overall effect size was small. In fact, most of the clustering seemed to occur based on passive stable factors — i.e., people with similar backgrounds or who lived relatively close to one another were more likely to form relationships.

** How well does the evidence for clustering in the kind of romantic partners we may choose hold up over time? Click here to continue to Part Two ...


Eastwick, Paul W.; Harden, K. Paige; Shukusky, Jennifer A.; Morgan, Taylor Anne; Joel, Samantha.   Consistency and inconsistency among romantic partners over time.  Journal of Personality and Social Psychology, Vol 112(6), Jun 2017, 838-859.

More Posts