Logistic mixed model parameter estimates (± *SE*) for the likelihood of passing by the confederate without deviating from path in relation to sex and height of the pedestrian, whether the confederate was of the same sex as the pedestrian, and their interaction (*N* = 1,056). Non-independence due to confederate ID was modelled as a random intercept.

^{a} Reference category is female

^{b} Reference category is ‘confederate of different sex as pedestrian’

^{c} Intercept at the level of confederate; variance estimate ± *SD*

^{d} (Pseudo-)*R*^{2}; see text for explanation.

Results from Study 3.