![[Adobe Stock]](https://sp-ao.shortpixel.ai/client/to_webp,q_glossy,ret_img,w_750,h_375/https://www.psypost.org/wp-content/uploads/2026/07/two-women-friends-talking-on-couch-750x375.jpg)
A new study published in the journal Social Psychological and Personality Science suggests that friends tend to have slightly similar personalities, but simply being alike does not actually predict how happy they are with their friendship. Instead, the research provides evidence that having friends who possess generally positive traits is much more important for relationship satisfaction. The findings suggest that how people perceive their friends’ personalities matters more to a friendship than strict personality matching.
“Our study found that friends were actually similar in personality, with the exception of extraversion,” said study author Hyewon Yang, who recently earned her doctorate in psychology at Michigan State University. “At the same time, people tended to perceive their friends as more similar to themselves than they actually were.”
She added that objective matching did not equal happiness. “Importantly, neither actual nor perceived personality similarity between friends predicted how satisfied they were with their friendships,” Yang said. “Rather, people’s own personalities and their perceptions of their friends’ personalities were more relevant to friendship satisfaction.”
People often assume that friends naturally gravitate toward one another because they share similar characteristics. This idea of social matching is a popular way to explain how human bonds form. Yet, friendship is a unique type of relationship because it is strictly voluntary and non-exclusive. People can choose their friends freely, making friendship an excellent context to study how much personality truly overlaps between companions.
“There has been a longstanding and interesting debate in the close relationships literature,” Yang told PsyPost. “Both theoretical accounts and laypeople’s beliefs suggest that people seek out similar others to affiliate with. However, empirical findings on whether such similarity is actually associated with better relationship well-being have been inconsistent.”
To measure these overlapping characteristics, psychologists often use the Big Five personality traits. This psychological framework breaks human personality down into five broad and stable categories. Extraversion describes how outgoing, energetic, and social someone tends to be. Agreeableness reflects a person’s capacity for kindness, empathy, and social cooperation.
Conscientiousness involves being organized, responsible, and goal-oriented in daily life. Neuroticism refers to a tendency to experience negative emotions like anxiety, worry, and sadness. Finally, openness involves curiosity, creativity, and a general preference for new experiences and abstract ideas.
Previous studies looking at these specific traits in friends have yielded mixed results. Past research often studied isolated pairs of friends rather than looking at larger friend groups. Older studies also sometimes mixed up actual similarity with perceived similarity. Clarifying the exact difference between these two concepts helps scientists understand the true role of personality in maintaining relationships.
Actual similarity measures how much two people genuinely share a trait based on their own separate self-reports. If both friends independently rate themselves as highly organized, they have high actual similarity in conscientiousness. Perceived similarity refers to a person’s personal belief that their friend shares their traits. This perceived overlap happens regardless of whether the friend actually rates themselves the same way.
To better understand these dynamics, the authors recruited 371 separate groups of exactly four friends. This resulted in a total sample size of 1,484 individual participants. The study took place online between the spring of 2023 and the spring of 2024. During the study, each group of four friends joined a muted video call together to verify their actual social connections.
The sample primarily consisted of young adults, with an average age of roughly 19 years old. Most participants were female, making up about 78 percent of the entire group. A majority of the sample identified as White or Asian. These participants had been friends for an average of about three and a half years.
Studying groups of four allowed the researchers to capture the complex nature of human social networks. Instead of just looking at two people in isolation, the study design mapped out a small community. This method captures how a single individual might rate multiple different friends. It also reveals how that same individual is perceived differently by various members of the group.
“Beyond the more common use of dyadic data to study similarity, to our knowledge, this is the first study to use round-robin data from friendship groups to examine these questions,” Yang said. Dyadic data refers to information collected from just a single pair of individuals. In contrast, a round-robin design involves every person in a group evaluating every other person in that same group.
Once their friendships were verified, participants completed an extensive online survey. Participants evaluated their own personalities using a standard questionnaire covering the Big Five traits. Because they were in groups of four, each participant also rated the personalities of the three other friends in their specific group. Afterward, participants answered questions about their overall satisfaction and happiness with each specific friendship.
“After considerable work with the statistical models, we developed an approach that accounts for the complex nested structure of friends embedded within dyads and groups,” Yang explained. “I hope this study can therefore also serve as a useful step for future researchers interested in examining similarity using round-robin designs.”
The researchers analyzed the data to see if friends actually shared traits and if they believed they shared traits. The analysis provided evidence for modest actual similarity among friends for four out of the five personality traits. Overall, friends were most alike in reality when it came to the trait of openness.
“The effect sizes were relatively small (e.g., rs = .05 to .10 for significant actual similarity),” Yang said. In statistics, the term “rs” refers to correlation coefficients, which measure how closely two variables are linked. An effect size in this range means the statistical connection is real but weak. “This suggests that although friends showed statistically significant personality similarity, personality may not be a domain in which particularly strong similarity among friends should be expected.”
Extraversion was the singular trait where friends did not show a statistically significant level of actual similarity. “I found it particularly interesting that we did not observe actual similarity in extraversion, that is, friends’ self-reported levels of extraversion were not significantly correlated,” Yang said. “This suggests that people likely do not have friends who are necessarily similarly extraverted or introverted.”
Openness might be the most shared trait because it heavily influences a person’s values and daily interests. Highly open individuals often seek out novel activities and intellectual conversations. People typically spend their free time doing activities they enjoy, and they naturally invite peers who enjoy those same activities. This shared preference for specific hobbies or environments tends to bring people with similar levels of openness together.
Beyond actual similarity, the data showed a strong tendency for participants to project their own traits onto their friends. Perceived similarity was significantly larger than actual similarity for almost every single personality trait. This suggests that people routinely assume their friends are more like them than they really are. Just as with actual similarity, openness generated the highest levels of perceived similarity among the friend groups.
The authors also tested whether actual or perceived personality matching predicted higher friendship satisfaction. They used complex statistical models to see if a perfect match between two friends’ scores led to a happier relationship. These mathematical techniques allowed the scientists to look for a specific spike in satisfaction that only occurs when both friends have the exact same level of a trait.
The analysis provided no evidence that similarity improved friendship satisfaction. Being exactly alike in personality did not reliably associate with how happy participants felt about their friends. This pattern held true whether the researchers looked at actual similarity or perceived similarity. Simply matching on a personality test does not guarantee a high-quality friendship.
Instead of similarity, the overall levels of certain positive personality traits drove friendship satisfaction. Participants who were more agreeable, conscientious, extraverted, and emotionally stable tended to report higher satisfaction with their friendships overall. The personal characteristics of the friends mattered greatly as well. Having friends who were viewed as highly agreeable and conscientious was consistently linked to better relationship well-being.
Emotionally stable friends are often easier to interact with on a daily basis. High levels of neuroticism can sometimes introduce unnecessary conflict or stress into social interactions. Because friendships are voluntary, people might have less patience for negative emotions than they would in family relationships. Having friends who remain calm and cooperative seems to foster a more peaceful and rewarding social environment.
Interestingly, how a person viewed their friend’s personality was a very strong predictor of satisfaction. A person’s subjective perception of their friend was more closely tied to friendship happiness than how the friend actually rated themselves. These findings suggest that seeing your friends as loyal, open, and agreeable is heavily linked to feeling content in the relationship.
It is important to consider the limitations and boundaries of this specific research project. “One important caveat concerns generalizability,” Yang said. “Our study focused specifically on personality similarity among friends, and our sample consisted largely of same-gender friends, especially women, who were relatively young and close to one another.”
Friendship dynamics might look very different in older populations or in different cultural settings. As people age, their social networks often shrink, which might change how personality traits influence friendship satisfaction over time. “There is considerable room for future research on other domains of similarity, such as moral values or political orientations, as well as other age groups, including midlife and older adults,” Yang added. “Patterns of actual and perceived similarity, and their implications for relationship well-being, may look different in these contexts.”
The cross-sectional nature of the data also prevents the researchers from knowing exactly how these friendships evolved. The findings represent just a single snapshot in time. Because of this, the authors cannot prove if similar personalities caused the friendships to form in the first place. They also cannot determine if friends slowly change their personalities to become more alike after spending years together.
Future research could follow friend groups over several years to observe how these bonds change. Tracking people over time might reveal if similar friends stay together longer or if dissimilar friends eventually drift apart. Scientists might also look at other domains of similarity outside of the Big Five personality traits.
“I have always been fascinated by (dis)similarity in close relationships: Are we actually similar to the people closest to us, and if so, is that similarity beneficial for our relationships, or not necessarily?” Yang said. “In a world where people increasingly seek out others who are similar to themselves or belong to the same so-called ‘in-groups,’ I have several ongoing projects aimed at unpacking these questions.”
Shared values, political ideologies, or religious beliefs might play a larger role in friendship satisfaction than basic personality traits do. Including these other factors could provide a more complete picture of what makes friendships thrive. “More broadly, I hope to better understand in which domains, and in what ways, being similar to or different from close others may contribute to satisfying and healthy relationships,” Yang concluded.
The study, “Friends’ Personality Similarity and Its Association with Friendship Well-Being,” was authored by Hyewon Yang, Atea Nelson, Lisa Stuckman, Grace Yancho, Lindsay S. Ackerman, M. Brent Donnellan, William J. Chopik, and Richard E. Lucas.