Carrots versus sticks: The relationship between training methods and dog-owner attachment

Abstract

The use of aversive-based training methods has been suggested to negatively affect dog-human attachment. However, the scientific evidence for this claim is relatively limited. Previous studies relied upon owner reports of training methods or on potentially confounded measures of attachment (e.g., eye gaze). The aim of the present study was to comprehensively and objectively investigate the relationship between aversive- and reward-based training methods and dog-owner attachment. Companion dogs (n = 34) recruited from 6 different dog training schools (3 reward-based and 3 aversive-based) were given a counterbalanced version of the Ainsworth Strange Situation Test. The presence and absence of the owner and a stranger in a room with the dog was manipulated over different episodes. Dogs' behavior was then analyzed for attachment-related behaviors: contact-maintenance, separation-distress and secure-base effect, as well as following upon separation and greeting upon reunion. Results showed no significant differences between groups for contact-maintenance and separation distress behaviors. However, dogs trained with reward-based methods, but not dogs trained with aversive-based methods, played more in the presence of the owner than in the presence of the stranger, and they also followed and greeted the owner more than the stranger, although these differences were found for only one procedure order. Our study is the first to investigate the relationship between training methods and attachment using a standard and well-validated method for the assessment of dog-owner attachment.