In line-ups of five models with varying factial expressions, Americans focused on the central figure while Japanese looked at background models. Credit: University of Alberta.