Comparing trial-and-error to errorless learning procedures in training pet dogs a visual discrimination

Abstract

Errorless learning is a stimulus fading procedure that can reduce incorrect responses in a stimulus discrimination task. Errorless learning has been described as a beneficial dog training procedure, although minimal research has been performed with dogs to support its use. The current study compared the effects of using an errorless learning to a trial-and-error learning procedure to train a stimulus discrimination in companion dogs. Eight dogs were trained, four in each group, with concurrent stimuli and either through gradually fading in an incorrect stimulus (errorless learning group) or by differentially reinforcing correct and incorrect responses (trial and error group). Correct responses, incorrect responses, frustration-related responses, and non-responses were measured and compared across the two groups. The results showed the errorless learning dogs were more likely to meet the criteria for learning the discrimination, showed faster speeds of acquisition, and displayed fewer frustration-related behaviours compared to the trial-and-error group. These results are discussed in relation to their benefits for training dogs more effectively and with better welfare during training, suggesting errorless learning could be more greatly utilized for canine training and research.