I have asked Dr Kate,a newly graduated PhD animal behaviourist, of Pets Behaving Badly, to tell us a little about her PhD findings. This will interest anyone concerned with shelter dogs and behaviour assessments…
Assessing shelter dog behaviour to determine adoption suitability: Meaningful or misleading?
By Dr Kate Mornement
Like many of you, I’m an animal lover and the high euthanasia rates of shelter dogs the world over is an issue that has always concerned me. Upon entering a shelter dogs typically undergo a behaviour assessment (or temperament test) to determine if they’re suitable for adoption. Problem behaviour, such as aggression towards people or other dogs, exhibited during these assessments or their shelter stay, is often used to justify euthanasia. I began my PhD with grand notions of saving all the shelter dogs. Over the course of my research, however, it became apparent that this was pipe dream albeit a worthy one! The very first thing a research student does is conduct a review of the scientific literature on their topic to find out what aspects have been researched and whether there are any gaps in the knowledge. I soon realised that very little was known about how shelters assess dog behaviour to determine adoption suitability and whether these assessments do what they are intended to do. To my shock, nobody had researched this in Australia. Woah!
The first part of my study involved an investigation of the beliefs and attitudes of the Australian public towards shelter dogs and shelter practices, particularly behaviour assessment, via a survey. We wanted to know whether Australians held positive or negative attitudes towards shelters and shelter dogs and were reluctant to adopt. Somewhat surprisingly, we found that Australians generally held positive beliefs and attitudes towards shelter dogs and were aware of shelter practices, including routine assessments of health and behaviour. Over 80% of our participants said they would obtain a future pet dog from an animal shelter or rescue organization. Despite this willingness to adopt a dog, one-third believed that adult shelter dogs often have behaviour problems. Another important finding from this part of the project was that almost 75% of potential adopters considered it important that methods used to evaluate adoption suitability are scientifically sound.
Previous research has questioned the reliability and validity of shelter assessments, so the next phase of my study was to review and describe current protocols used by Australian animal shelters. This involved travelling to nine shelters in five states. I also wanted to find out whether those responsible for assessing dogs were confident in their current test and in their ability to accurately assess dogs. The results revealed that, although Australian shelters do their best with available resources, the protocols we reviewed lacked standardization in their content and methodology and none had been evaluated scientifically to establish whether they actually worked. Despite this, the shelter staff we interviewed unanimously agreed it was important that such tests are scientifically valid. In addition, staff responsible for assessing dogs lacked confidence in their ability to accurately assess dog behaviour and in the protocol they currently used. Even those staff members with extensive experience lacked confidence perhaps indicating a greater appreciation of the complexity of canine behaviour and the deficits in current assessment protocols. One-third of the staff members we interviewed had not received training in assessing dog behaviour and, of those that had, most reported having received informal on-the-job training.
To address these issues we developed the Behavioural Assessment for Re-homing K9’s (B.A.R.K.) protocol tested the it in a working shelter. The standardized B.A.R.K. protocol was comprised of the most common subtests already used in the tests we reviewed. We wanted to know if B.A.R.K. yielded similar results when the same dog was assessed by the same person on two different occasions (called test-retest reliability) and if two experienced assessors get similar results when assessing the same dog (inter-rater reliability). We also wanted to know if B.A.R.K. could predict behaviour in the new home (predictive validity) and examined this via a post adoption questionnaire. The results showed that the inter-rater reliability was high across the five behavioral traits measured. However, the test-retest reliability was relatively weak. This is concerning given the assessments took place just 24 hours apart, with the same experienced assessor in the same environment. Additional research is needed to address this by examining time effects and previous exposure on the behavioural responses given by dogs to the B.A.R.K. protocol.
Finally, we evaluated the predictive validity of the B.A.R.K. protocol, which was also quite weak. Although the results did support its usefulness as a tool for predicting fearful and friendly behaviour, B.A.R.K. was not effective in predicting aggression or other problem behaviour post adoption. Other studies have found similar results and this could be due to problems inherent in assessing behaviour in a highly stressful environment, where a stable human-canine relationship is absent. Stress is known to effect cognition and the expression of behaviour which may be unstable following the trauma of capture, confinement, loss of familiar social companions and control over the environment and the novelty of the shelter environment. Therefore, behaviour exhibited by shelter dogs during an assessment may not be indicative of its normal behaviour under ordinary circumstances, such as in a familiar home environment. Exclusively relying on such tools to inform decisions regarding adoption suitability and euthanasia is, therefore, problematic and not recommended.
My research has brought to light the many challenges facing shelters and behavioural scientists trying to ensure that methods used to evaluate pet potential are meaningful and accurate. Despite these challenges, ongoing research in this area is critical to safeguard the welfare of dogs in the shelter system and the wider community. Foster care programs have recently emerged as an alternative to long shelter stays and warrant further study. Assessments conducted in a less stressful home environment may be more predictive of future behaviour.
About Dr Kate
Dr Kate Mornement is a PhD qualified animal behaviourist, consulting pet owners and industry through her business Pets Behaving Badly. You can follow her on Facebook and Twitter. For more information about Kate’s research and her services please see her website.