Revising and validating the random search model for competitive search

A random search model was fit to a total of 2592 visual search times on a single-target detection task. By using a competing homogeneous background and uniform stimulus material, specifying viewing distance, controlling the presentation of search task material, and eliminating some options for extreme search strategies, very high correlation coefficients were found when a random search model was fit to both the individual data and to pooled data. A response time parameter was incorporated into the traditional random-search model and very good predictions of search performance were obtained.