Effects of contingent and noncontingent reinforcement on the emotional behavior of domestic dogs (Canis familiaris)

Abstract

Intermittent schedules of reinforcement are frequently used in dog training. During fixed-time (FT) schedules, reinforcement delivery is time-based and response-independent. In contrast, during fixed-interval (FI) schedules, a response produces a reinforcer if a certain amount of time has elapsed since the last reinforcer was delivered (i.e., FI schedules are time-based and response-dependent). FT schedules are typically used in noncontingent-reinforcement interventions with the aim of decreasing an undesired behavior. Although the effectiveness of these interventions has been reliably shown with human and nonhuman animal learners, little is known about the emotional behavior that is related to these contingencies. In the present study, the target behavior (stepping on a mat) of six dogs was initially trained (shaped) and maintained on a FI 15s schedule. In addition, dogs were exposed to a FT 15s schedule, a combination of FT 15s and FI 15s schedules, and an extinction-only condition (no reinforcement was delivered for the target response). We found that rates of emotional responses (e.g., lip/nose licking, vocalizing) varied in prevalence across dogs and phases. Although lip/nose licking was displayed in high rates across all dogs and phases, it was generally lower in baseline and extinction phases. This suggests that these phases produced less frustration-related behavior. The effect-size calculations for dogs engaging in pacing, sneezing, and vocalizing underscored the varied prevalence of the behaviors across phases and dogs and yielded mostly small- to medium-size effects. These results are especially relevant for the training of companion and working dogs and their well-being during training sessions. Future research should continue investigating emotional behavior in dogs across various schedules of reinforcement (e.g., variable schedules).