The Theoretical Background of the Similarity Hypothesis
The similarity hypothesis is one of the most robust findings in research on interpersonal attraction, and it gives scientific support to the everyday intuition that birds of a feather flock together. Donn Byrne's work on attitude similarity, published in 1971, laid the foundation of the field. Byrne used an experimental method known as the bogus stranger paradigm, in which participants were shown the attitudes of a fictitious other person, and demonstrated repeatedly that the more closely those attitudes resembled the participant's own, the greater the liking for that person. The effect has been confirmed not only for attitudes but across a variety of dimensions, including personality traits, values, interests, socioeconomic background, and level of physical attractiveness. Several mechanisms have been proposed to explain why similarity generates attraction: consensual validation, in which the other person confirms that one's own attitudes are valid; the predictability of interaction with someone who resembles oneself; and the reduction of cognitive load that comes with it.
Personality Similarity and Relationship Satisfaction
In the context of the Big Five, a substantial body of research has accumulated on the relation between personality similarity and relationship satisfaction. Meta-analyses find that personality similarity between partners does in fact exist, but that the effect size is moderate and not equivalent across all five factors. Similarity in agreeableness and conscientiousness shows a comparatively consistent positive association with relationship satisfaction. The likely reason is that when levels of cooperativeness and of responsibility are aligned, everyday friction decreases. Similarity in extraversion and openness, by contrast, has a weaker and more situation-dependent association with satisfaction. For neuroticism the crucial thing is not similarity but that both partners be low: where both are high, relationship quality declines even though the two match each other closely. The similarity hypothesis therefore applies differently factor by factor, and it cannot be reduced to a blanket rule that more alike is always better.
The Limits of the Similarity Hypothesis and Its Integration With Complementarity
The similarity hypothesis is a powerful theory, but it is not universal. First, the effect of similarity changes with the stage of a relationship. Similarity exerts a strong effect during the initial formation of attraction, while in long-term relationships other factors gain in importance, among them communication skills, commitment, and shared experience. Second, there is no need for similarity along every dimension; the finding is that similarity in core values matters most. Third, an integrative perspective that incorporates the complementarity hypothesis is required. On the dimension of dominance and submissiveness, complementarity, with one partner leading and the other following, can contribute to how smoothly the relationship runs. Research on compatibility turns on that distinction: which dimensions call for similarity, and which ones benefit from complementarity. So the sheer number of things you have in common is not a score in itself. It helps more to separate the areas where being alike does the work, such as core values, from the areas where covering for each other keeps things moving, such as who takes the lead.