International Journal of Comparative Psychology

Adult humans typically overestimate the number of items in regularly arranged stimulus sets compared to randomly arranged sets. This regular-random numerosity illusion (RRNI) was examined comparatively in adult humans, chimpanzees, and rhesus monkeys. Neither nonhuman primate species showed evidence of the illusion when trained to pick the larger of two sets of randomly arranged dots on a computer screen and then shown regularly arranged sets. Adult humans, given the same task and instructed to select the larger set, showed the illusion, although there were individual differences. Experiment 1 used somewhat different methodologies with the human participants compared to the nonhuman animals, but Experiment 2 presented the identical method to naïve human participants and naïve rhesus monkeys with minimized training, equal exposure to the different arrangement types, and very limited instructions to human participants. In that situation, human and monkey performance was very similar and reflected the RRNI. These results demonstrate that nonhuman animals also are susceptible to the RRNI, and they also indicate how methodological differences used during training both within and between species can impact results of comparative assessments.