This 2022 study by Hammond et al. reinforces a growing body of evidence demonstrating that breed is not a reliable predictor of agonistic behavior toward people, that is, behaviors such as barking, growling, or biting. By combining two established behavioral assessment tools (the Dog Impulsivity Assessment Scale (DIAS) and the Positive and Negative Activation Scale (PANAS)) and analyzing data from 1,845 dog owners across 25 breeds, the researchers found no meaningful behavioral differences between dogs of breeds commonly targeted by breed-specific legislation and those that are not. While minor statistical differences emerged at the group level, the wide variation within each breed rendered these differences irrelevant when considering any individual dog. The study adds weight to the argument that breed-based policies lack scientific justification, and it underscores that responsible dog guardianship, more than breed, influences behavior outcomes. As with previous studies using different methodologies, these findings call into question the rational basis of laws and practices that attempt to regulate dog ownership based on appearance or breed labels.
We include this study because it combines two very different assessment tools regarding correlations between dog breeds and agonistic behavior from those used in previous research on this question. Hammond and colleagues’ findings confirm those of previous studies (Ott et al, 2008; Schalke et al, 2008, Creedon and Ó Súilleabháin, 2017), but do so more comprehensively in the number and choice of breeds included. These combined diagnostic instruments yielded a null finding when comparing groups of dogs who are members of breeds commonly subject to restricted ownership with breeds that are not restricted. Such restrictions are justified by legislators’ opinions that members of some breeds are more likely to engage in warning and biting behavior toward people than dogs of other breeds. The researchers found this not to be the case, The second major finding was that in breed to breed comparisons, while some statistically, but not necessarily clinically significant differences were found on the population level, the variability within breeds was such that these findings had no significance for the likelihood of agonistic behavior for an individual dog. This replicates findings from various other studies, but looked at through a different lens, confirming that breed is not an important factor in predicting agonistic behavior in dogs (Morrill, 2022).
The researchers distributed an online survey that combined 2 existing behavior surveys designed to capture owners’ reports of their dogs’ behavior. The survey was distributed to various websites and social media likely to be used by owners of pedigreed dogs, including the UK kennel club and subsidiaries and also directly emailed to prominent breeders to request their participation. This contact method may have mitigated some of the risk of breed mislabeling by owners in online surveys. They looked only at dogs in the UK to avoid cultural and other confounds. They received enough responses to include those from owners of dogs of 25 different, non-closely related breeds. Among the 8 breeds they grouped as “legislated,” the legislation exists outside the UK, since only a few breeds are restricted in the UK. Questionnaires were analyzed from UK and Republic of Ireland owners of dogs who are restricted in other countries, along with those for dogs from 17 other breeds who are not subject to restrictions anywhere, for a total of 1,845 owner responses.
The authors provide an overview of the reliability and validity assessments that have been don on the instruments used–the DIAS (Dog Impulsivity Assessment Scale) and PANAS (Positive and Negative Activation Scale). Both have been shown to have acceptable internal consistency and some test-retest reliability, along with some convergence with other measures of diagnostic or predictive validity. However, it is important to remember that since no “gold standard” psychometric instrument for dog behavior that meets the general standards of reliability and validity for human psychometric instruments has yet been developed and validated, this is always a limitation on such surveys when used to assess canine behavior. That being said, the strength of the findings here may be in the fact that these instruments are so different from ones previously used to address the question of warning and biting behavior prevalence by breed. The DIAS is, in fact, intended as a measure of impulsivity, which has been shown to have implications for aggression in human behavior, but is not itself designed to predict threatening behavior in dogs. This means that the similarity between the findings here and those of other population level studies of such behaviors in dogs is particularly striking with regard to commonly legislated breeds (Duffy et al, 2008).
In summary, the importance of these findings is confirmatory. They focus specifically on comparing a variety of dogs from breeds that are not legislated against to ones from breeds that are and finds no differences between the groups, once again demonstrating that no improvement in public safety can be expected from restricting dog ownership or husbandry by breed.
References
Creedon, N. and Ó Súilleabháin, P.S., 2017. Dog bite injuries to humans and the use of breed-specific legislation: a comparison of bites from legislated and non-legislated dog breeds. Irish veterinary journal, 70, pp.1-9.
Duffy, D.L., Hsu, Y. and Serpell, J.A., 2008. Breed differences in canine aggression. Applied Animal Behaviour Science, 114(3-4), pp.441-460.
Ott, S.A., Schalke, E., von Gaertner, A.M. and Hackbarth, H., 2008. Is there a difference? Comparison of golden retrievers and dogs affected by breed-specific legislation regarding aggressive behavior. Journal of Veterinary Behavior, 3(3), pp.134-140.
Schalke, E., Ott, S.A., von Gaertner, A.M., Hackbarth, H. and Mittmann, A., 2008. Is breed-specific legislation justified? Study of the results of the temperament test of Lower Saxony. Journal of Veterinary Behavior, 3(3), pp.97-103.
