A rating-only approach to perceptual discrimination

Abstract Confidence ratings have been used in the absence of a binary decision to attenuate bias effects on decision criteria, thus, mitigating a common source of error. We tested whether a rating-only approach could be used to effectively discriminate whether a dynamic grid consisted of predominantly blue or predominantly orange pixels, and target present from target absent trials in a visual search task. Experiment 1 (n = 42) confirmed that participants could use the rating-only approach to discriminate between stimulus types. Experiment 2 (n = 99) tested whether the rating-only approach could discriminate luggage images that did or did not contain a weapon, and whether using ratings could mitigate the low prevalence effect (where rarely occurring targets often go undetected). The rating-only approach discriminated target present from target absent trials, however the absence of clear low prevalence effects precluded an investigation of the use of a ratings-based approach for countering this cause of error.