A new study published in the Journal of Research in Personality suggests that feeling satisfied in a romantic relationship depends heavily on specific combinations of how you view yourself and how you view your partner. Rather than a simple rule of opposites attracting or birds of a feather flocking together, the research indicates that matching political views tend to benefit couples, while perceiving a partner as kinder than oneself produces the highest relationship quality. These patterns provide evidence that romantic success relies on different psychological rules depending on the specific characteristics being evaluated and the cultural background of the couple.
Marta Kowal, a researcher at the Institute of Psychology at the University of Wrocลaw in Poland, designed this research to address a persistent puzzle in relationship science. Past studies have produced mixed results regarding whether romantic partners need to be similar to be happy.
“The saying goes that opposites attract. However, scientific evidence largely suggests otherwise, people who are more similar tend to have more satisfying relationships,” Kowal said. “At the same time, some studies have reported the opposite pattern.”
Kowal noticed that previous studies heavily relied on people from Western, Educated, Industrialized, Rich, and Democratic societies, often referred to as WEIRD populations. This lack of diversity left a gap in understanding how universal these romantic preferences might be.
“Moreover, much of what we know comes from WEIRD samples,” Kowal explained. “Therefore, utilizing a large dataset of 41,606 individuals from 74 countries provided a great opportunity to examine when similarity matters, when partner idealization matters, whether these associations depend on the specific trait being considered, and, finally, whether these patterns are consistent across cultures.”
To explore these dynamics, the researchers analyzed data from tens of thousands of individuals currently in romantic relationships. These participants completed online surveys translated into their local languages, answering questions about their relationship status and the length of their partnership.
The survey asked people to rate themselves and their romantic partners on nine specific traits. These traits included health, kindness, physical attractiveness, religiousness, resources, social class, education, political orientation, and age. The participants used simple numbered scales, such as an eleven-point scale for health and attractiveness, to evaluate both themselves and their significant others.
To measure relationship quality, the study used established psychological questionnaires. Participants completed a short version of the Triangular Love Scale, which breaks love down into intimacy, passion, and commitment. They also filled out a relationship satisfaction scale to report how happy they were with their partnership overall.
Kowal used a specialized statistical technique to compare the self-ratings and partner-ratings simultaneously. This mathematical approach allowed the researchers to see what happens when partners are perfectly matched compared to when one person is rated higher than the other.
The data showed that kindness and physical attractiveness were the strongest predictors of a high-quality relationship. For these socially desirable traits, a concept known as partner idealization emerged as highly important.
“The effect sizes varied considerably across traits,” Kowal told PsyPost. “Kindness was the strongest predictor: participants’ self-ratings and their ratings of their partners’ kindness together accounted for approximately 21% of the variance in relationship satisfaction, which is quite substantial for a single trait assessed with a single item.”
Participants reported the highest levels of love and satisfaction when they believed their partner was kinder and more attractive than they were. “Physical attractiveness was similarly strong,” Kowal added, while noting that value-based and demographic traits showed more modest effects overall.
“Perhaps the most intuitive takeaway is that for qualities that are universally valued (like kindness and physical attractiveness), what seems to matter most is not that you and your partner are evenly matched, but that you perceive your partner as very kind and attractive, ideally even more so than yourself,” Kowal explained.
Additionally, these positive traits showed an additive effect, meaning that relationship quality was highest when individuals rated both themselves and their partner very highly. This means that a baseline of high mutual kindness forms a strong foundation for romance.
“Importantly, this partner idealization effect goes hand in hand with a strong ‘more is better’ pattern, relationship quality is highest when both partners are perceived as high on these traits,” Kowal said.
The pattern looked completely different for value-based traits like political orientation. For political beliefs, strict similarity was the strongest predictor of a happy relationship.
“For values-based traits, however (particularly political orientation), similarity does matter in a more symmetrical way: the further apart people perceive themselves and their partners to be, the lower their relationship quality tends to be, regardless of which direction the gap falls in,” Kowal said.
Discrepancies in political views predicted lower relationship quality regardless of which partner leaned further to the left or the right. “So ‘opposites attract’ appears to be mostly a myth, but the reasons why depend a great deal on what dimension we are talking about,” Kowal noted.
The researchers expected to find significant differences between how men and women evaluate partner traits, but the results showed largely consistent patterns across genders. Past evolutionary theories often suggested that men and women prioritize very different qualities in a mate, making these results quite notable.
“I was actually quite surprised by small gender differences,” Kowal said. “Given the long-standing debates in evolutionary psychology about sex differences in mate preferences, I expected to find more pronounced asymmetries.”
Kowal expected that a man’s relationship quality would be more strongly related to a partner’s physical attractiveness than a woman’s. While some minor differences emerged, they were not the defining feature of the data.
“While there were a few domain-specific gender differences (e.g., men reported higher passion when partnered with younger partners, and higher relationship satisfaction when they perceived themselves as more right-wing than their partners), these were the exception rather than the rule,” Kowal said. “In general, men and women were rather in agreement.”
Culture, on the other hand, played a major role in shaping what people value in a partner. Kowal categorized countries based on several cultural indicators, including human development, gender equality, individualism, and relational mobility. Relational mobility refers to how much freedom individuals have in a society to choose or leave their romantic partners.
“Another important finding was the presence of cross-cultural differences,” Kowal noted. “Specifically, the pattern of associations across the three trait domains varied across the 74 countries. Cultural characteristics (including the Human Development Index (HDI), individualism, gender equality, and relational mobility) moderated many of the observed associations.”
In highly modernized countries with high relational mobility and individualism, socially desirable traits like kindness and attractiveness had the strongest impact on relationship happiness. In less modernized countries, the rules shifted toward status and demographic traits, meaning that matching on social class and educational background was a stronger predictor of relationship stability.
While the study features an impressive global sample, the reliance on self-reported data means the results capture subjective perceptions rather than objective reality. Because the data only includes one partner’s perspective, it measures perceived similarity rather than actual similarity.
“The most important limitation is that this study relies on individual-level, self-reported perceptions rather than dyadic data from both partners,” Kowal explained. “This means that what we are capturing is perceived similarity and perceived partner traits, not actual similarity or objectively assessed traits.”
A person deeply in love might view their partner through rose-colored glasses, rating them higher simply because they feel happy. This psychological habit makes it tricky to determine the exact cause and effect.
“That is not necessarily a weakness (perceived similarity often predicts relationship quality better than actual similarity does), but it does mean we cannot rule out that relationship quality itself is partly driving the pattern,” Kowal said. “People who are in love tend to perceive their partners more favorably, which could inflate the observed partner-idealization effects.”
Future research should involve surveying both members of a couple over a longer period of time. This step would help determine if actual similarity matters as much as perceived similarity. Tracking couples over the years would also help explain how these perceptions change as relationships grow older and face new challenges.
The study, “Partner idealization and perceived partner similarity predict relationship quality across 74 countries,” was authored by Marta Kowal.