Abstract
An increasing number of studies have investigated the visual perception of social interactions. A commonly used paradigm in this research is the comparison of facing and non-facing dyads in visual search, where facing dyads are found faster than non-facing dyads. This has commonly been interpreted as evidence of a specialized system for social interaction processing. In a series of experiments (N=40 each), we use this same paradigm to first replicate previous findings and then show that similar effects can be found when presenting non-social pairs of arrows. We then exclude the possibility of different causes producing these similar results by using combined face-arrow pairs, which again produced a response time advantage for pairs that directionally cue each other. These findings indicate that the search advantage found for facing dyads is a product of the directional cues present within these arrangements, not the fact that they are processed as social interactions per se. This highlights the need for existing paradigms used in this area of research to be further refined as this common manipulation not only changes the perception of social interaction but also how participants attend to these displays.