Exploring a framework for matching to sample with visual stimuli in dogs (Canis familiaris)

Abstract

Dogs have been studied for their olfactory abilities, yet their capacity for same-modal visual matching to stimuli remains underexplored. Research focused on scent detection, and cross-modal odour matching. However, little attention was given to visual matchings. This study investigated dogs’ ability to match visual stimuli. Two monitors were used: one displaying a single image, and the other divided into four quadrants. Four dogs were trained to select the stimuli on the second monitor as displayed on the first. In the discrimination task, an errorless learning approach was implemented, where the correct stimuli were presented at full brightness while the incorrect shape was dimmed. As the stages progressed, the incorrect shape’s brightness increased. All dogs successfully matched to sample at low brightness, and they were able to perform better than chance. However, they didn’t meet the criteria to discriminate between shapes when both were presented at full brightness. After discontinuing errorless learning, the dogs went under a pure conditional discrimination test. Almost all dogs performed better than chance, while there were some significant findings specific to each dog. The findings contribute to understanding canine visual perception and contribute to a methodological framework for visual-visual matching tasks. Future studies may refine these methods to enhance canine performance in discrimination learning.