Skip to main content

Verified by Psychology Today


Do Opposites Attract?... It Depends on How They Interact

Research finds that actions speak louder than personality.


We’ve all heard how “opposites attract." But we're also told that “birds of a feather flock together."

The fact that both of these adages have been passed down for so long suggests that the role of similarity in relationships is not a simple matter.

Most research indicates that we do prefer to affiliate with others who are similar to us—who share our values and interests.1,2,3 But some claim that when it comes to personality traits, we may be most interested in complementarity. This means that for some traits, similarity is most desirable, but for others, we prefer someone who is our opposite.

The type of complementarity that has received the most attention from researchers examines two traits: affiliation (warm and friendly vs. cold and hostile) and control (dominant vs. submissive).

According to this theory, we will prefer someone who is similar to us on affiliation (warm people like other warm people, and cold people like other cold people) and opposite on dominance (dominant people pair off with submissive people).4

On the other hand, we might expect that everyone, regardless of their own personality, would prefer positive traits in others. For example, even cold people should still prefer to be with someone who is warm.

Only a few studies have examined complementarity in existing romantic relationships: In one study, in which undergraduate students rated their own parents’ personalities, complementarity appeared quite often, but greater complementarity was actually seen in divorced than still-married couples.5 Another study of romantic couples found very little complementarity, except within highly satisfied couples.6 Yet another study found good evidence for the phenomenon, with couples being similar on affiliation and opposite on dominance.7 Finally, a study of same-sex female couples found no similarity on affiliation and found that those who had opposite personalities when it came to dominance were actually less satisfied with their relationships.8

As can be plainly seen, these results are quite inconsistent.

Trying to determine a couple’s compatibility or relationship satisfaction based on general personality traits is difficult: It's why personality-based match-making algorithms don’t work! Research points to the idea that where traits really matter is how they’re expressed during our actual interactions with partners.

Despite our general levels of traits like warmth or dominance, we regularly alter how warmly or dominantly we behave in different situations.9 Jenny Cundiff and colleagues sought to solve this problem by examining the extent to which couples expressed these traits while actually interacting with their spouses, instead of focusing on general personality traits.10

In two studies, Cundiff and colleagues asked married couples to come into the lab to have different types of discussions. These couples rated the affiliation (warmth-coldness) and control (dominance-submissiveness) of their spouses during that interaction, along with their own levels of anxiety, anger, and relationship satisfaction after the discussion. In the first study, couples engaged in a positive interaction (taking 1-minute turns describing their partners’ positive characteristics), a negative interaction (describing the partner's negative characteristics), or a neutral interaction (describe partners’ daily schedule), followed by a discussion about an issue on which they disagreed. In the second study, the couples either discussed a disagreement or collaborated on a task (working together to plan the best route and schedule for running a list of errands, using a map of a hypothetical town).

Both studies found strong evidence for similarity in affiliation, regardless of the type of interaction. That is, if one spouse was warm, so was the other, and if one spouse was cold, so was the other. But the picture for control was more complicated. In positive or collaborative interactions, complementarity of control occurred—one spouse led and the other followed. But for negative interactions or disagreement interactions, the opposite was true—both spouses were controlling or both submissive. In those more negatively toned interactions, it is likely that the two spouses were actually competing for control, rather than one allowing the other to lead.

The results also showed that participants felt better after interactions in which their spouses expressed positive traits. High levels of control and low levels of affiliation were associated with spouses reporting more anger and anxiety during the disagreement discussion, as well as lower relationship satisfaction. That is, people responded more positively to the interaction when they felt that a spouse's behavior was warm and submissive. There were also some interesting gender differences in these reactions. Wives were the least anxious and angry when they themselves were low in affiliation and their husbands were high in affiliation (that is, the wives were relatively cold, but the husbands were warm). Husbands were the most angry and the least satisfied with their relationships when both spouses were low in affiliation. Wives, on the other hand, were most satisfied with their relationships when both spouses were low in control (that is, both submissive). Even though complementarity occurred in certain types of interactions, it didn’t lead spouses to feel any more satisfied with those interactions—what really mattered was the extent to which positive traits were expressed.

These results suggest that when it comes to personality traits, it is neither the case that “birds of a feather flock together,” nor that “opposites attract." Rather, the answer is different for the traits of warmth and dominance, and, more specifically, what really matters is how much those traits are actually expressed as we discuss both positive and negative issues with our partners. And when it comes to what makes for a more satisfying interaction, showing positive traits is more important than showing traits that complement your partner’s.

Gwendolyn Seidman, Ph.D. is an associate professor of psychology at Albright College, who studies relationships and cyberpsychology. Follow her on Twitter for updates about social psychology, relationships, and online behavior. Read more more articles by Dr. Seidman on Close Encounters.


1 Newcomb, T. M. (1961). The acquaintance process. New York: Holt, Rinehart & Winston.

2 Byrne, D., & Nelson, D. (1965). Attraction as a linear function of proportion of positive reinforcements. Journal of Personality and Social Psychology, 1, 659-663.

3 Lewis, K., Kaufman, J., Gonzalez, M., Wimmer, A.,m A Christakis, N. (2008). Tastes, ties, and time: a new social network dataset using Social Networks, 30, 330-342.

4 Kiesler, D. J. (1983). The 1982 interpersonal circle: A taxonomy for complementarity in human transactions. Psychological Review, 90, 185-214.

5 Tracey, T. J. G., Ryan, J. M., & Jaschik-Herman, B. (2001). Complementarity of interpersonal circumplex traits. Personality and Social Psychology Bulletin, 27, 786-797.

6 Markey, P. M., & Markey, C. N. (2007). Romantic ideals, romantic obtainment, and relationship experiences: The complementarity of interpersonal traits among romantic partners. Journal of Social and Personal Relationships, 24, 517-533.

7 Smith, T. W., Traupman, E. K., Uchino, B. N., & Berg, C. A. (2010). Interpersonal circumplex descriptions of psychosocial risk factors for physical illness: Application to hostility, neuroticism, and marital adjustment. Journal of Personality, 78, 1011-1036.

8 Markey, P. M., & Markey, C. N. (2013). The complementarity of behavioral styles among female same-gendered romantic couples. Personal Relationships, 20, 170-183.

9 Buss, A.R. (1979). The trait-situation controversy and the concept of interaction. Personality and Social Psychology Bulletin, 5, 191-195.

10 Cundiff, J. M., Smith, T. W., Butner, J., Critchfield, K. L., & Nealey-Moore, J. (2015). Affiliation and control in marital interaction: Interpersonal complementarity is present but is not associated with affect or relationship quality. Personality and Social Psychology Bulletin, 41, 35–51.