Similarity Principle of Attraction

Marian M Morry. Encyclopedia of Human Relationships. Editor: Harry T Reis & Susan Sprecher. Sage Publications. 2009. 

Many individuals hold beliefs about the importance of similarity for attraction. These beliefs can be expressed in sayings such as “Birds of a feather flock together” and “Opposites attract.” This entry focuses on beliefs expressed by the first of these sayings—that individuals prefer others who are similar to themselves. One of the most basic principles of interpersonal attraction is that individuals form relationships with people who are similar to themselves and dissolve relationships because of a loss of similarity. This has been referred to as the similarity-attraction hypothesis. These effects are found across relationship types from interactions among strangers to friendships and romantic relationships. Therefore, attraction in this context refers to a variety of variables from liking or willingness to interact with a stranger to loving or relationship quality in a marriage. Research on this topic has been conducted in a number of areas, such as whether similarity leads to attraction or whether attraction leads to similarity, whether it is actual or perceived similarity that is related to attraction, and why similarity is important. This entry examines each of these topics.

Similarity Leads to Attraction

Beginning with Aristotle’s essay on friendship in 330 BCE, the idea that similarity increases interpersonal attraction has permeated our beliefs. Near the end of the 19th century, Sir Francis Galton provided empirical verification of this belief when he found that spouses were similar in many attributes. However, his findings could not address the cause-and-effect relation between similarity and attraction. In the 1950s and 1960s, Theodore Newcomb conducted one of the earliest studies examining the influence of similarity on attraction at the University of Michigan. He examined friendship formation among male housemates who were initially strangers. At first, men’s friendships were influenced mostly by perceived similarity, but as the semester continued and the men learned more about their housemates, these friendships were influenced mostly by actual similarity. In both cases, however, higher similarity was associated with higher attraction. This research suggested that similarity leads to attraction, but the naturalistic setting and lack of experimental manipulation and control still raised questions about cause and effect.

To examine the cause-and-effect relation in the similarity-attraction hypothesis, Donn Byrne and his colleagues conducted a series of studies using the “bogus stranger paradigm.” In the bogus stranger paradigm, a participant is led to believe that there is another individual in the study. However, the other individual does not actually exist—hence, the “bogus stranger.” In these types of studies, attitudinal similarity of the bogus stranger to the participant is experimentally manipulated to determine its effects on attraction. Attraction in these studies is typically measured by a question about how much the participant liked the stranger or how willing the participant would be to work with the stranger on a future task. These studies, then, asked whether an individual would be more likely to express a desire to work with an unknown stranger (attraction) the more similar the stranger’s attitudes were to one’s own attitudes (similarity). Supporting the cause-and-effect relation, individuals who were randomly assigned to the experimental condition in which the bogus other was presented as being similar to the individual were more attracted to the person than individuals who were randomly assigned to condition where the bogus other was presented as being dissimilar. These findings are robust, with the importance of similarity on attraction extending from interactions with strangers to marital relationships.

Arthur Bochner and Michael Sunnafrank have criticized the similarity-attraction research, mostly that of Donn Byrne’s early work, indicating that this research deals with stranger or bogus-stranger relationships, those with unilateral awareness and no interdependence, and confuses the timing of learning about someone’s attitude and meeting the person. Therefore, they criticized whether the attraction-similarity findings can be generalized to the initiation and development of interpersonal relationships. However, research examining whether actual or perceived similarity is the important predictor of attraction supports the similarity-attraction relation in real, interdependent relationships. This issue is discussed in more detail later. Although in real life we may not know another person’s attitudes before we meet them, we tend to meet people in situations that allow us to infer these attitudes, such as through our involvement in school, work, social clubs, or church. Sharing similar interests in these activities (e.g., a social club) implies that the two individuals will also share other beliefs or interests. That is, we can infer similarities in a number of domains based on similarities in one domain. As is discussed next, these inferences may or may not be accurate.

Despite that Donn Byrne’s research indicated a linear relation between similarity and attraction, this relation depends on a variety of other factors, including the kinds of similarity being shared and the perceiver’s personality. In terms of the types of similarity being shared, attitude similarity is more important than personality similarity, and value similarity is more important than demographic similarity for attraction. The effects of similarity on attraction are also stronger for central than peripheral attitudes. Central attitudes refer to issues that are important to an individual’s self-view or self-concept, whereas peripheral attitudes would have little or no importance to the self. The similarity-attraction relation is also greater for attitudes with high, as opposed to low, heritability and for personality traits that are more important (e.g., caring, jealous) compared with less important (e.g., quiet, cold) for the relationship. Finally, the similarity effect is stronger for positive traits than for negative traits.

Personality characteristics also influence preferences for the type of similarity important to attraction. For example, low self-monitors prefer attitudinal similarity over activity similarity, whereas high self-monitors prefer activity similarity over attitudinal similarity. Individuals high in social anxiety indicate greater attraction to agreeing strangers and lower attraction to disagreeing strangers than do individuals low in social anxiety. Similarly, individuals with a positive self-concept, as measured by the self-assurance scale, prefer similar to dissimilar others, whereas individuals with a negative self-concept do not show this preference for similarity.

Attraction Leads to Similarity

Although the similarity-attraction research indicates that these variables are related and among strangers similarity leads to attraction, in ongoing relationships, the direction of influence may differ. Arthur Bochner notes that relational partners assume that they should have a variety of things in common and communicate to foster this impression. In fact, he argues that one purpose of communication is to foster perceptions of similarity and create an impression of being an interesting person. Supporting this role of communication, attitude alignment or changing one’s own attitude to match the attitude of a partner is found when individuals are aware of an attitudinal discrepancy (the individuals have discussed it), when the issue is central to the partner (they know what each other believes and what is important to them), and when the interactions are among dating partners rather than strangers.

Perceptions of similarity in ongoing relationships may also occur because individuals assume similarity and, therefore, project themselves onto their partners. Although we believe we hold accurate impressions of our friends and romantic partners, partner impressions are affected by a variety of factors; for instance, observer’s own behaviors and attitudes influence their ratings of others’ behaviors and attitudes. Thus, we are attracted to people who are similar to us, the similarity-attraction hypothesis, but may also perceive similarities that may not be there. Marian Morry and her colleagues have referred to this latter aspect as the attraction-similarity hypothesis and suggest that an ongoing relationship attraction to another person may lead individuals to perceive self-other similarities. Supporting the attraction-similarity hypothesis, Marian Morry and her colleagues have found that asking individuals to write about the most positive or negative event in their relationship caused these individuals to rate their relationship satisfaction (a measure of attraction) as higher or lower, respectively. These changes in attraction then caused changes in perceived similarity in both friendships and dating relationships. That is, increasing ratings of attraction increased ratings of perceived similarity and decreasing ratings of attraction decreased ratings of perceived similarity. In summary, it appears that in ongoing relationships, attraction influences perceived similarity.

Actual or Perceived Similarity

Having found that similarity is important to attraction, and vice versa, researchers are also interested in whether actual similarity is necessary or whether perceived similarity is sufficient. Supporting the effects of actual similarity, a number of studies found evidence of assortative mating or the tendency of like to marry like (e.g., sharing the same religion). In fact, most individuals become friends with or marry people who are similar to the self. Evidence for assortative mating has been found for intelligence, social class, religion, race, sensation seeking, and attachment styles, among others. Close relationship partners exhibit greater than chance similarity on these variables, and this actual similarity is positively associated with attraction and the continuation of the relationship. Because most of these variables predate the onset of a relationship, these findings indicate that in interdependent relationships, actual similarity leads to attraction.

Considerable discrepancies, however, exist between self and peer ratings of attitudes, values, beliefs, and personality, suggesting that perceptions and reality are not the same. For example, individuals perceive their partners to be more similar to themselves than the partner self-rates in terms of personality traits such as extraversion, neuroticism, openness, agreeableness, and conscientiousness, and interpersonal traits such as being kind, affectionate, controlling, or dominant. These perceptions of similarity then may impact the similarity-attraction relation. For example, perceived, not actual, value similarity predicts liking among college roommates. In romantic relationships, perceived not actual similarity of coping styles is the best predictor of satisfaction.

In the context of ongoing relationships, the question of actual or perceived similarity refers to whether one is accurate in one’s perception of the partner. That is, does Peter’s perception of Sue match Sue’s self-ratings? To test for actual or perceived similarity, researchers ask both individuals to rate themselves and their partner on a variety of traits, opinions, behaviors, and so on. It is then possible to test whether there is actual similarity. Peter and Sue rate themselves the same; a perception of similarity, Peter rates Sue as being similar to his own self-ratings, but Sue does not rate herself similar to how Peter rates himself; or a combination of the two, some similarity in self-ratings and Peter rates Sue as being more similar to himself than is warranted. A variety of research studies indicates that, although somewhat accurate in their perceptions of friends, dating partners, and spouses, individuals tend to perceive greater similarity than warranted. That is, although there is some actual similarity, individuals project themselves onto their intimate others and perceive similarities that are not there to the same degree. This tendency to perceive greater similarity than warranted has been found for attitudes, personality characteristics, attachment styles, behaviors, and so on. In addition, these perceptions of similarity are found across a variety of relationships from friendships to dating relationships to marriages. As noted earlier, in ongoing relationships, experimental changes in relationship satisfaction, by writing about a positive or negative event in the relationship, caused changes in perceived similarity. These results also support the importance of perceived rather than actual similarity.

Explanations for Why Similarity and Attraction Are Related

Another area of research in the similarity principle of attraction domain is why these variables are related. Explanations have included rewards of interacting, balance theory, and general expectations. The first explanation, rewards of interacting, indicates that individuals are motivated to be logical and accurate in interpreting their environment. Similarity with others validates their view of the world and confirms that they are correct in their thinking. In addition, an interaction with someone who shares your attitudes, beliefs, or values results in fewer disagreements and conflicts, which should foster a sense of safety. Similarity can also facilitate interaction as each individual can predict the other’s responses.

Fritz Heider’s balance theory has also been proposed as an explanation for the relation between similarity and attraction. According to this theory, individuals prefer relationships that are balanced. For example, the participant (P) likes his or her friend (O), and the participant has a particular attitude, behavior, or trait (X). With two positive relations (PO and PX), the third relation should also be positive (e.g., the friend should have or be perceived to have X) to achieve a balanced relationship. Supporting the current application of balance theory, people perceive similarities between themselves and intimate others based on projections of the self. Theodore Newcomb hypothesized that increases in attraction (PO) or of the relevance of the characteristic (PX) to the relationship would motivate individuals to change the relationship into a balanced one. Consistent with his reasoning, manipulations of attraction influence perceptions of similarity, individuals overestimate attitude similarity more in close relationships such as marriage than in the stranger paradigm, and similarity effects on attraction are stronger for central than peripheral attitudes and for moderate compared with low relationship-relevant traits.

Another explanation for the relation between similarity and attraction is that individuals expect similar others to like them. That is, similarity implies liking, and we are attracted to individuals who like us (i.e., reciprocity of liking). In line with this reasoning, John Condon and William Crano manipulated both how positive a stranger evaluated the participant and similarity of the stranger to the participant. To manipulate the stranger’s evaluation, they indicated that the stranger thought that the participant was above average in a number of ways and expected to like him or her and wanted to work with him or her or provided no such information. Both positivity and similarity affected the participant’s attraction to the stranger. In addition, similarity led to inferred positive evaluations by the stranger, and these inferred evaluations then led to attraction. This expectation of liking explanation is also consistent with the role of communication in developing relationships. More specifically, individuals tend to reciprocate the level of disclosure. Disclosure provides the opportunity to learn about similarities, which implies the other likes the self, and increases liking for the other.

Current and Future Directions

Current research continues to look at the types of similarity important for attraction, the direction of influence between similarity and attraction, and the reasons that similarity and attraction are related. For example, Eva Klohnen and her colleagues examined attitudes and personality traits and found substantial actual similarity on attitudes, but not on personality characteristics. However, greater perceived similarity on personality characteristics, and especially attachment styles, was the best predictor of relationship satisfaction (attraction). In addition, this perceived similarity-satisfaction relation continued even when actual similarity was controlled for. Together these results suggest that the type of similarity being examined is important and that perceived similarity may be more important to attraction than actual similarity.

Research on the similarity principle of attraction can also be enhanced by using more diverse research methods. For example, research on the direction of influence can be enhanced with techniques from speed-dating research. For instance, in speed-dating research, it would be possible to test whether initial actual similarities in the interaction leads to later liking or whether liking at the end of the speed-dating session leads to later perceived similarity. Longitudinal research could also be used to determine when the effects of similarity on attraction among strangers become the effects of attraction on perceived similarity in ongoing relationships.

Finally, research is still needed to explore the reasons for the similarity and attraction relation. As noted earlier, these explanations involve rewards of interacting, balance theory, and expectations. The motivational and cognitive underpinnings of these explanations, however, need to be explored. Such questions could include: How do individuals develop expectations of similarity? Do these expectations vary across relationship types (friendships, marriages)? Do they change over time or with experience?


Although there are a variety of variables that are important for interpersonal attraction, one such variable that has captured the attention of both lay people and researchers is similarity. A number of studies have indicated a positive, linear relation between similarity and attraction. The strength of this relation varies based on the importance of the characteristics, one’s personality traits, and so on. In addition, both perceived similarity and actual similarity are important correlates of attraction. The direction of influence also seems to vary based on the closeness of the relationship, with actual or perceived similarity leading to attraction among strangers, but attraction leading to perceived similarity in ongoing relationship.