Human directed aggression in domestic dogs (Canis familiaris): Occurrence in different contexts and risk factors.

To best understand this article in the context of the literature on growling, snarling, snapping, and biting behavior (incidence and correlates), please see National Canine Research Council’s complete analysis here.
Article Citation: 
Casey, R. A., Loftus, B., Bolster, C., Richards, G. J., & Blackwell, E. J. (2014). Human directed aggression in domestic dogs (Canis familiaris): Occurrence in different contexts and risk factors. Applied Animal Behavior Science, 152, 52-63.

National Canine Research Council Summary and Analysis:

This U.K. study is included as it is an attempt to compare warning and biting behavior in various contexts, to determine whether a dog is more or less likely to express these behaviors in multiple contexts. Casey et al. (2014) studied barking, lunging, growling, and biting (as a group of behaviors labelled “aggression”) in three different contexts: unfamiliar people entering the home, unfamiliar people outside of the home, and household members. No distinction was made between warning behaviors and actual biting, or between injurious and non-injurious bites. Despite this broad, all-encompassing definition of aggression, owners reported low levels of such behavior in each context (approximately 3%, 7%, and 5% expressed the behaviors of interest towards family members, unfamiliar persons entering the home, and unfamiliar persons outside the home, respectively), in this survey of owners’ self-reports, recalled over the dog’s lifetime. The incidence findings are particularly surprisingly since a bark is not necessarily a threat behavior; dogs often bark from excitement or happiness which might have been expected to increase the incidence reported.

Sampling bias is a recurring problem in the dog bite and aggression literature, and this study is no exception, relying for 50% of the results on people attending or participating in dog shows; and the 27% response rate demonstrated a significant self-selection bias. These sampling problems may at least partially explain the dramatic difference between these results and the much higher incidence rates found in Guy et al. (2001a), where the sample was a more plausibly representative group of veterinary clients and the response rate much higher. The large sample size (N=3,897) can be seen as strength, but it’s compromised by the low return rate only 26.8% of the distributed surveys (N=14,566).

The authors also reported the surprising result of low measures of agreement between fearful behaviors (e.g., hiding or running away) and barking, lunging, growling, and biting. The central finding was a lack of congruence between dogs who barked, lunged at, and bit strangers in the home as compared to those who threatened strangers outside the home or family members in any context. The authors concluded from this lack of consistency in behaviors between contexts that neither a breed nor an individual dog is “dangerous,” “vicious,” or “safe” as a characteristic. Instead, they concluded that learning that occurs in one context but not another may play a large determining role in fearful or threatening behavior.

Abstract and Link to Purchase Full Text of the Original Article: