Fig. 1

An example of a best-worst scaling (BWS) trial. Participants view a subset of the faces to be rated, and select the “best” (in this case, most attractive) and “worst” (in this case, least attractive) from the subset. This “best”/“worst” decision is easy to understand, naturalistic and relies only on the faces presented in the current trial, with no need to remember previous responses. These faces, from the Face Research Lab London Set (DeBruine & Jones, 2017), are for illustration purposes only, and were not used in the studies reported here