Research overview: false-positive detections in occupancy modeling

Standard occupancy models assume that detections are never mistaken, but in practice species can be misidentified, a particular risk when automated classifiers or volunteers confuse similar-looking cats, or when sign surveys misattribute tracks. This overview surveys false-positive occupancy models that relax the no-error assumption, jointly estimating false-positive and false-negative detection rates to avoid the inflated occupancy and distorted habitat relationships that misidentification causes. It discusses how confirmation of a subset of detections, or multiple detection methods, supplies the information needed to separate true from false records. As machine-learning classification of camera-trap images becomes routine, accounting for misclassification grows increasingly important. The overview connects this statistical correction to the practical reality of imperfect species identification in camera and sign-based felid surveys.