The Psychology of Attraction

In brief

For most of history, attraction was treated as mysterious, a matter of chemistry or fate. From about 1950, social psychologists began to measure it, and a stubborn structure appeared. Three factors predict, better than chance, whom a person will befriend or desire. The first is proximity: we like the people we happen to be near, an effect so mechanical it can be read off a floor plan. The second is similarity: we are drawn to those whose attitudes, values, and backgrounds resemble our own, and the folk belief that opposites attract is close to false. The third is physical attractiveness, which shapes first impressions more powerfully than people say it does, while also dragging along a halo of unearned assumptions about character. None of the three is the whole story, but each is real, replicated, and, taken together, they turn a supposed mystery into something a researcher can forecast.

The full treatment

The problem it answers

Common sense offers contradictory maxims: birds of a feather flock together, but opposites attract; love is blind, but beauty matters. Because both halves of each pair feel true, folk wisdom explains everything and predicts nothing. The psychology of attraction set out to replace proverbs with measured effects: hold a situation constant, vary one factor, and see whether liking moves. What emerged is not a theory of the soul but a set of reliable regularities about the ordinary conditions under which one person comes to like another.

Proximity, and the study that fixed it

The founding demonstration is Leon Festinger, Stanley Schachter, and Kurt Back's study of a student-housing complex, published as Social Pressures in Informal Groups (1950). They surveyed residents of Westgate and Westgate West, apartment blocks built for married MIT veterans after the war, where people had been assigned to units largely by availability rather than choice. Asked whom they saw socially, residents named their nearest neighbors far more often than distant ones. The striking part was how fine-grained it was. On a given floor, people who lived one door apart were much likelier to become friends than people three or four doors apart, and the effect did not track raw physical distance so much as "functional distance," the number of times a layout forced paths to cross. Residents whose doors faced the stairways or mailboxes, where everyone passed, had more friends across the whole building. Where you happened to be placed, not who you were, predicted whom you came to like.

Robert Zajonc gave the effect a mechanism in 1968 with the mere-exposure effect: repeated exposure to a stimulus, whether a face, a nonsense word, or a Chinese character, increases liking for it, even when the exposures are too brief to be consciously recalled. Proximity works, on this account, because nearness manufactures repeated exposure, and familiarity, within limits, breeds liking rather than contempt.

Similarity, and the bogus stranger

The second pillar is similarity. Donn Byrne built a research program around it, summarized in The Attraction Paradigm (1971), using a method he called the "bogus stranger." A participant fills out an attitude survey, then is shown the survey of a stranger who does not exist, secretly constructed to agree with the participant on some fixed proportion of items. Across many studies, attraction to the stranger rose in an almost linear relationship with the proportion of shared attitudes. The similarity that matters is broad: attitudes and values, but also personality, background, education, and even physical attractiveness, where people tend to pair with partners of roughly their own level. The popular idea that opposites attract has found little support; where complementarity appears, it is usually within a frame of deep underlying similarity.

Physical attractiveness, and the halo it casts

The third pillar is the one people most resist admitting. In 1966, Elaine Walster and colleagues ran a "computer dance" at the University of Minnesota: over seven hundred incoming students were told a computer had matched them to a date, when in fact partners were assigned at random. The researchers had discreetly rated each student's physical attractiveness beforehand and measured how much each liked their date and wanted to see them again. Of everything measured, including intelligence, personality, and social skill, only one thing predicted liking: the partner's physical attractiveness. Nothing else came close, for men or women.

Attractiveness carries a further payload. Karen Dion, Ellen Berscheid, and Elaine Walster (1972) coined the phrase "what is beautiful is good": shown photographs, people assumed the attractive strangers were kinder, more interesting, more successful, and better married than the plain ones. This physical-attractiveness stereotype is a halo effect, a first impression on one dimension bleeding into judgments on unrelated ones, and it has real consequences in hiring, sentencing, and teaching.

A distinction worth keeping

Later work sharpened the similarity finding into a distinction that matters. What predicts attraction, especially past a first meeting, may be perceived similarity, the belief that another person resembles you, rather than actual, measured similarity. In an influential meta-analysis, R. Matthew Montoya and colleagues (2008) found that actual similarity predicts attraction mainly in brief laboratory encounters, while in real, ongoing relationships it is the perception of being alike that carries the weight. The effect is real; its cause is subtler than the bogus-stranger studies implied.

Lineage

The study of attraction is a branch of the experimental social psychology that grew from the field theory of Kurt Lewin, under whom both Festinger and Schachter trained; the housing study is Lewin's environment-shapes-behavior program applied to friendship. Its ultimate root is the founding of psychology as an experimental science by Wilhelm Wundt, who insisted that mental life could be measured rather than merely introspected. The similarity work descends from the reinforcement tradition, Byrne treating agreement as a reward. The attractiveness work fed into evolutionary psychology, which reinterpreted beauty preferences (symmetry, averageness, cues of health) as evolved mate-selection heuristics, and into the modern science of close relationships built by Berscheid, Walster, and later Eli Finkel.

The strongest case for it

The findings are robust in the way that matters most: they replicate, across decades, samples, and methods. Proximity has been confirmed in dormitories, police academies (where alphabetical seating, not choice, predicted friendship), and workplaces. Mere exposure is one of the most reliably reproduced effects in all of psychology. The similarity effect appears whether measured by survey, by observed pairing, or by meta-analysis. And the primacy of physical attractiveness in first encounters has survived every attempt to explain it away. The program delivered exactly what it promised: it took a domain everyone thought was governed by ineffable chemistry and showed that a large share of the variance is predictable from a handful of ordinary variables, several of them external to the people involved.

The strongest case against it

The objections are serious and come from within the field.

First, self-report and behavior diverge sharply. People say they want particular traits in a partner, but those stated preferences predict little about whom they are actually drawn to. Paul Eastwick and Eli Finkel, in a 2008 speed-dating study and later work, found the standard sex difference in stated preferences (men prioritizing looks, women earning power) but showed that these self-reports failed to predict who was actually attracted to whom once people met face to face. What predicts attraction in the moment is not the trait profile a person endorses in the abstract.

Second, the ecological validity of the classics is limited. The proximity study captured a specific population (young married veterans in assigned housing in 1950) and platonic friendship, not romantic love, in an era before cars, phones, and later the internet loosened the grip of physical nearness. Whether functional distance still governs attraction when much social life is mediated by screens is an open empirical question, not a settled one.

Third, the similarity effect was oversold in its original form. As Montoya and colleagues showed, the tidy linear relationship of the bogus-stranger studies reflects an artificial setting; in real relationships, actual similarity does far less work than the paradigm suggested, and much of the apparent effect is perceived similarity, which can be a consequence of liking as much as a cause.

Fourth, the generalizability of the attractiveness stereotype is contested. Cross-cultural work finds that the specific content of "what is beautiful is good" varies: collectivist cultures tend to attribute the traits their own norms prize (in one line of research, integrity and concern for others rather than dominance) to attractive faces. The halo is general; its contents are local.

Where it stands now

The three predictors survive, but as pieces of a larger and more dynamic picture rather than as the whole of it. Proximity and mere exposure are treated as established; similarity is treated as real but heavily mediated by perception; physical attractiveness is accepted as a powerful driver of initial attraction whose influence recedes as people come to know one another. The frontier has moved to what happens after the first meeting, the domain of relationship science, where Finkel and others argue that compatibility is co-constructed over time and poorly forecast by any static list of traits, a claim that reframes the entire dating-algorithm industry's premise. The old finding, that attraction is not random, holds. The newer finding is that it is not simple either.

Test yourself

Think of the people you count as close. How many did you meet because you were assigned to sit, live, or work near them, rather than because you sought them out? If the honest answer is most of them, notice how thoroughly a factor you never chose, sheer proximity, has shaped a part of your life you probably think of as freely and deeply your own.