Details of the samples of cats and dogs tested.

A long-standing question in the study of quantity discrimination is what stimulus properties are controlling choice. While some species have been found to do it based on the total amount of stimuli and without using numerical information, others prefer numeric rather than any continuous magnitude. Here, we tested cats, dogs, and humans using a simple two-way spontaneous choice paradigm (involving food for the first two, images for the latter) to see whether numerosity or total surface area has a greater influence on their decision. We found that cats showed preference for the larger amount of food when the ratio between the stimuli was 0.5, but not when it was 0.67; dogs did not differentiate between stimuli presenting the two options (smaller vs. larger amount of food) regardless of the ratio between them, but humans did so almost perfectly. When faced with two stimuli of the same area but different shapes, dogs and humans exhibited a preference for certain shapes, particularly the circle, while cats’ choices seemed to be at chance level. Furthermore, cats’ and dogs’ reaction times were equal across conditions, while humans were quicker when choosing between stimuli in trials where the shape was the same, but the surface area was different, and even more so when asked to choose between two differently sized circle shapes. Results suggest that there is no universal rule regarding how to process quantity, but rather that quantity estimation seems to be tied to the ecological context of each species. Future work should focus on testing quantity estimation in different contexts and different sources of motivation.