Friendship Formation and Development

Beverley Fehr. Encyclopedia of Human Relationships. Editor: Harry T Reis & Susan Sprecher. 2009. Sage Publication.

In the classic film Casablanca, the saloon owner delivers a famous line: “Louis, I think this is the beginning of a beautiful friendship.” This entry addresses questions surrounding the beginnings of friendships. How do people “make” friends? What is the process by which an acquaintanceship is transformed into a friendship? Research reveals that friendship formation is a complex process in which a number of factors must converge. First, the environment must bring two people into contact with one another. Second, the situation must be “right” for a friendship to develop. For example, both people must be at a point in life where they have the time and resources to devote to a new friendship. The qualities that people possess also play a role in friendship formation; individual factors such as attractiveness and social skills matter. Finally, friendship is ultimately a dyadic process. In other words, it takes two to form a friendship. As will be seen, friendships are more likely to form when the two people share important similarities, when liking is mutual, and when self-disclosures are reciprocated.

The Environment

For a friendship to develop, two people must be brought into contact with one another. The role of physical proximity in friendship formation is well documented. For example, in a classic study, married students living in a student housing complex were asked to name the three people in the complex with whom they socialized most. Two thirds of the people named lived in the same building, and two thirds of these people lived on the same floor. Other studies have shown that people are likely to form friendships with those who live nearby (i.e., residential proximity). Proximity effects also have been shown in the workplace, in college dormitories, and in classrooms.

What is it about neighborhood, workplace, and school settings that promotes friendship formation? The short answer is that these settings provide opportunities for contact. The greater the amount of contact between two people, the greater the likelihood that they will become friends. However, physical proximity is becoming less important than it was in the past. Many people are now relying on the Internet as a venue for meeting potential friends. Thus, it is possible that in the future, environmental factors will be less crucial for friendship formation, although the people with whom we rub shoulders on a day-to-day basis will probably continue to be candidates for friendship formation.

The Situation

A number of “chance” factors influence whether or not friendships develop. One such factor is whether the two individuals will have opportunities for ongoing interactions and whether they will be able to interact on frequent basis. Importantly, both people also must be “available” for this kind of relationship.

Opportunities for Interaction

When two people meet each other, they usually know whether this is likely to be a one-time occurrence (e.g., chatting with a fellow passenger on an airplane) or whether their interactions will be ongoing (e.g., chatting with a new coworker). We are more likely to pursue a friendship with a person when we believe that there will be future opportunities to interact with him or her. This was demonstrated in a classic study in which research participants watched a videotape of three people having a discussion. Some participants were led to believe that they would not be meeting anyone on the videotape; others were told that they would have a one-time meeting with one of the people on the videotape; still others were told that they would be meeting with one of the people during the next 5-week. Those who expected to meet during a five-week period rated the person in the videotape the most positively, followed by those who expected a one-time meeting. Those who did not expect to meet at all provided the least positive ratings. Other studies have shown that when we expect to interact with someone over an extended period, we tend to emphasize the positives and downplay the negatives so that our future interactions with the person will be smooth and enjoyable.

Frequency of Interactions

As just discussed, if we anticipate future interactions with a person, we evaluate him or her more positively than if we do not. Does the frequency of those interactions matter? The answer is yes. Considerable research shows that the more often we see someone—or even a photograph of a person—the more we like him or her. This phenomenon is referred to as the mere exposure effect. There is one exception: If we initially dislike someone, repeated contact can cause us to like the person even less. In general, however, the more contact we have with a person, the more we will like him or her, and the greater the probability that a friendship will form.


We do not form a friendship with every person with whom we have ongoing interactions. Another factor must be considered, namely whether we have room in our lives for a new friendship. Friendships require a number of resources—time, energy, and even money. Other commitments in life, such as time-consuming studies, work demands, and existing relationships (romantic partner, family, friendships) can prevent us from pursuing a promising new friendship. A friendship can form only if each person is available for this kind of relationship.

Thus, circumstantial factors affect the development of friendships. Friendships are more likely to form if two people expect that they will have ongoing interactions, that they will be able to see each other on a frequent basis, and when each person has the time and energy to devote to forming a new relationship.

The Individual

Even if the situation is “right” for the development of a friendship, there is no guarantee that a friendship will form. Another important class of variables comes into play, namely whether the other person has qualities that we want in a friend—and vice versa. A number of characteristics make a person a desirable friendship candidate, including physical attractiveness, social skills, and responsiveness.

Physical Attractiveness

It is well known that looks are important in determining attraction to potential romantic partners. However, physical attractiveness also matters in the friendship selection process. Research conducted with adults and children shows that physical attractiveness is correlated with popularity—the better looking someone is, the greater the likelihood that he or she will be sought out as a friend.

Why are good-looking people at an advantage when it comes to making friends? One reason is the tendency to assume that “what is beautiful is good.” In other words, when people are attractive on the outside, we assume that they also are attractive on the inside and attribute positive qualities to them. Research also has shown that we assume that people who are good looking are similar to us in personality and attitudes. (As is discussed later, it is well-established that we are attracted to similar others.) Finally, evidence also indicates that people who are physically attractive may have better social skills than do those who are less attractive. From childhood on, good-looking people experience positive reactions from others, which results in increased self-confidence and social competence. Thus, interactions with physically attractive people may actually be more enjoyable than are interactions with those who are less attractive. There are a number of reasons, then, why good-looking people are pursued as friends.

Social Skills

It has been said that making friends is a skilled performance, much like learning a new sport or learning to drive a car. Research conducted with adults and children confirms that people who have good social skills are liked more than are those who are less socially skilled. Social skills include being competent at initiating conversations, asking appropriate questions, and showing interest in what the other person is saying. Social skills also include nonverbal behaviors such as appropriate patterning of eye contact and gaze and following norms for interpersonal spacing (e.g., respecting the other’s personal space). Social skills are crucial at the beginning stage of friendships; those who are socially skilled are better at getting friendships “off the ground.” Once a friendship is established, however, it becomes less important to be socially skilled and more important to be competent in providing warmth and support.


Another individual-level characteristic that is closely related to social skill competence is responsiveness. A responsive individual pays attention to questions he or she is asked and makes appropriate, relevant responses. These behaviors convey interest, liking, and concern. As a result, the interaction partner feels more comfortable opening up, which, as we shall see, is an important element of the friendship formation process (see section on Self-Disclosure). Indeed, several experiments have shown that when people are interacting with a responsive (versus a nonresponsive) interaction partner, they feel liked by him or her, they report greater liking for him or her, and they see the person as someone who potentially could become a friend.

Thus, several individual characteristics are associated with friendship formation. Those who are physically attractive, who have good social skills, and who are responsive are likely to be sought out as friends.

The Dyad

A friendship is a relationship between two people. Thus, analyses of friendship formation must consider the characteristics of each individual, as well as the interplay or the “chemistry” between them. As discussed next, friendships are most likely to form when liking is reciprocal, when self-disclosure is mutual, and when the two people share similarities.

Reciprocal Liking

“How I like to be liked, and what I do to be liked!” These words, penned by the 19th-century English writer Charles Lamb, are as applicable today as they were 200 years ago. In a classic demonstration of this phenomenon, groups of same-sex strangers engaged in weekly discussions during a 6-week period. Before the first meeting, each participant was told that based on personality information gathered earlier, the researchers could predict which group members would like him or her. (The names of these group members were actually randomly selected.) As expected, participants expressed the greatest liking for those group members who they believed liked them.

Interestingly, the perception that another person likes us may cause us to behave in ways that confirm that expectation. In another landmark study, researchers led participants to believe that their interaction partner either liked or disliked them. Those who believed their partner liked them engaged in more intimate self-disclosure, were more pleasant, and demonstrated fewer distancing behaviors than did those who believed they were disliked. Importantly, these behaviors lead the interaction partner to like them. Thus, when another person likes us, we tend to like them in return. Even the belief that another person likes us creates liking because it puts in motion a self-fulfilling prophecy whereby we behave in ways that produce the liking that we initially expected.


We are generally attracted to a person who self-discloses to us because revealing personal information sends a message that he or she likes us and desires a closer relationship with us. Indeed, many studies have demonstrated that we like people who engage in intimate self-disclosure more than we like those who engage in non-intimate disclosures. The one exception to this pattern is when another person reveals “too much too soon”—highly intimate self-disclosures from a stranger can elicit dislike, rather than liking.

We also like those to whom we have self-disclosed. The effect of engaging in self-disclosure (rather than being on the receiving end) was examined in a study in which pairs of strangers participated in a 10-minute “get acquainted” discussion. The more a participant self-disclosed, the more he or she liked the other and saw the other person as a potential friend. Thus, in general, the greater another person’s self-disclosure, the more we like him or her. We also like those to whom we have self-disclosed.

At the early stages of relationships, it is important for disclosures to be reciprocal. If Person A reveals something intimate about himself or herself, Person B needs to reciprocate with an equally intimate disclosure. Indeed, considerable evidence indicates that reciprocity of disclosure is associated with greater liking for an interaction partner. Reciprocity is considered important in establishing trust in a relationship. When we first meet another person, we do not know whether he or she can be trusted to keep our self-disclosures in confidence, whether he or she might use the information we have disclosed to hurt us, or whether he or she might ultimately reject us. We are more willing to risk being vulnerable if the other person is also taking the same risk. Thus, self-disclosure generally takes the form of “turn taking” in which we reveal personal information and then assess whether the other person reciprocates and whether he or she can be trusted with the information that we have shared. If the other person seems trustworthy, we will gradually increase the intimacy of our self-disclosures, while monitoring whether the other person is also increasing the intimacy of his or her disclosures. Once trust is established, it is not necessary for each self-disclosure to be reciprocated in each specific interaction; rather, there is an assumption of reciprocity over the long term.


One of the most widely researched predictors of friendship formation is similarity. A broad base of evidence indicates that people are likely to become friends with those who are similar to them in demographic characteristics (e.g., age, physical health, education, religion, family background), residential proximity, social status, physical attractiveness, and so on. Although adult friendships are the focus of this literature, most of these effects have been obtained in studies with children and adolescents as well.

The classic domain in which similarity effects have been investigated is attitude similarity. In early investigations of the role of similarity in attraction, participants were asked to complete a questionnaire assessing their attitudes on a variety of issues (e.g., politics, religion, tuition increases). They were then shown the same questionnaire supposedly filled out by a person who would be their partner during an experiment. In actuality, the researchers completed the questionnaire so that it appeared either similar to that of the research participant or dissimilar. Participants then were asked to give their impression of this person, before meeting him or her. Those who believed that the partner held attitudes similar to their own reported greater attraction and liking than did those who believed that they and their partner held dissimilar attitudes. Subsequent research demonstrated that these findings also apply to real-world friendships, namely that people tend to form and maintain friendships with those who are similar to them in attitudes. People also tend to develop friendships with those who share their values.

Similarity effects also are pronounced for activity preferences. We are more likely to form friendships with people who enjoy the same hobbies, sports, and leisure preferences that we enjoy. Interestingly, there is little evidence that people become friends because of personality similarity, although similarity effects have been found for more relationally oriented characteristics such as social and communication skills.

There is one domain in which similarity effects are found for children’s and adolescents’ friendships, but not for adults’ friendships, namely similarity in prosocial and antisocial behaviors. These effects are strongest for antisocial behavior. For example, research has shown that aggressive children seek out other aggressive children as early as preschool and that this tendency becomes more pronounced with age. Research also shows that adolescents tend to seek out as friends those who are similar to them in drug and alcohol use and school delinquency (e.g., cutting classes, quitting school).

Overall, substantial evidence indicates that we are likely to become friends with those who are similar to us. The only area in which similarity effects seem to be weak or nonexistent is personality similarity. Thus, it seems to matter less that our friends share our characteristics, than that they share our attitudes, values, social competencies, and leisure preferences.

Why are we more likely to form friendships with similar, rather than dissimilar, others? The most common explanation is that our views are validated by interacting with someone who shares them. Put another way, we feel more confident that we are “right” in our thinking if we encounter someone else who thinks just like us. Another explanation focuses on the enjoyment of interactions. The idea is that interactions are smoother and more pleasant if we agree with another person on most things. Disagreement tends to make interactions tense and strained.

In summary, a number of dyadic factors promote the formation of friendships. The two people must like each other. They must engage in a process of mutual self-disclosure in which the intimacy of information revealed gradually increases over time. And, finally, potential friends should be similar in most ways.