Document Type


Degree Name

Master of Arts (MA)




Faculty of Science

First Advisor

Christian Jordan

Advisor Role

Thesis Supervisor


The relationship between physical similarity and seating preference was examined using two observational studies and one laboratory study. Using Campbell et al.’s (1966) seating adjacency formula, Study 1 found significant aggregation by glasses-wearing status and sex when observing seating patterns in a library computer lab. That is, men sat beside other men, glasses-wearers by other glasses wearers, and so on. Study 2 broadened this methodology by examining a wider variety of physical traits in university classrooms; specifically, race, sex, glasses-wearing, hair length and hair colour. Broadly speaking, multivariate tests revealed an overall tendency for people to sit beside physically similar others more frequently than expected by chance alone. These results remained significant even when controlling for sex, race or prior acquaintanceship. Study 3 conceptually replicated these results in a laboratory setting. Photos of participants were coded for physical similarity to a confederate and attractiveness by independent coders. The more physically similar participants were to a confederate, the closer they sat. This finding remained significant even when controlling for sex, race and attractiveness similarity. Other nonverbal measures were also examined as potential dependant variables; however, only speech disfluencies emerged as a significant correlate with physical similarity. As physical similarity increased, the number of speech disfluencies uttered by the participant during a short interaction also increased. The potential moderating role of implicit self esteem and body esteem on the physical similarity/seating distance relationship was also examined. However, these results were nonsignificant. Finally, as perceived similarity to the confederate increased, so did positive ratings of the confederate using Likert scales; however, perceived similarity did not predict any nonverbal measures. The current research rules out simple matching on sex, race or attractiveness as potential explanations for this finding. An evolutionary kinrecognition mechanism is discussed as a potential mechanism behind these findings, drawing on Debruine’s (2004a) work. However, much work is left to be done to more concretely determine the driving force behind this relationship.

Convocation Year