Elsevier

Applied Animal Behaviour Science

Volume 151, February 2014, Pages 75-83
Applied Animal Behaviour Science

Development of the behavioural assessment for re-homing K9's (B.A.R.K.) protocol

https://doi.org/10.1016/j.applanim.2013.11.008Get rights and content

Abstract

Globally, millions of dogs enter the welfare shelter system each year. Before being made available for adoption dogs are typically screened for their suitability as companions by way of a “temperament test” or behavioural assessment. In Australia, the majority of dogs that fail their behavioural assessment are subsequently euthanased. Previous research has identified a lack of standardisation, in both the content and methodology, and a lack of scientific validation of such screening procedures. This poses a significant welfare concern for shelter dogs; life and death decisions might be made based on invalid assessments of behaviour. The aim in this study was to develop a standardised shelter dog behaviour assessment, called the behavioural assessment for re-homing K9's (B.A.R.K.) protocol, implement it into an operational animal shelter and evaluate the reliability and predictive validity of the tool. The B.A.R.K. protocol consists of 12 subtests that aim to imitate everyday situations a pet dog is likely to encounter. The behavioural trait with the highest overall, and statistically significant (n = 48, P < 0.01), inter-rater reliability was ‘fear’, with a mean correlation of 0.95 across all B.A.R.K. subtests. ‘Fear’ also had the highest overall, and statistically significant (n = 46, P < 0.01), test–retest reliability with a correlation of 0.82 across all B.A.R.K. subtests. The overall inter-rater reliability of the B.A.R.K. protocol was moderate to strong however the test–retest reliability was relatively weak. Amongst dogs that initially passed the test and were subsequently rehomed, the predictive validity of the protocol was also quite poor, with ‘fear’ (r = 0.42, n = 67, P < 0.01) and ‘friendliness’ (r = 0.49, n = 67, P < 0.01) being the only measures that proved to be predictive. The results of the study imply that a standardised behavioural test may be of less value in identifying the suitability of dogs for placement in the community than is currently believed. If so, this has significant implications for how such tests are employed.

Introduction

A number of tests have been developed to assess dogs’ aptitude for a variety of different purposes. These include working roles such as police dogs (Slabbert and Odendaal, 1999), military dogs (Haverbeke et al., 2009) and guide dogs for the blind (Serpell and Hsu, 2001). They also include assessment of dogs’ suitability as pets (Lucidi et al., 2005). Despite the depth and breadth of canine behavioural testing, there remains a lack of standardisation in the way such tests are implemented (Diederich and Giffroy, 2006), especially in animal shelters where the implications of using invalid protocols are significant.

Prior to being made available for adoption, shelter dogs in Australia and elsewhere typically undergo a behavioural assessment or “temperament test”. Information collected during the assessment, in which dogs are exposed to a series of stimuli intended to replicate real world situations, is used to identify stable behavioural tendencies in order to predict future behaviour that may be expressed in similar circumstances (Dowling-Guyer et al., 2011). The purpose of the assessment is to determine whether a dog will make a suitable companion for potential adopters (Christensen et al., 2006) and, more recently, to optimise the match between adopter and dog (Bollen and Horowitz, 2008). Dogs that pass the assessment are deemed “adoptable” and made available for sale. Those that fail, usually because of chronic health or behavioural problems, are euthanased or undergo rehabilitation, being subsequently reassessed at a later date.

Approximately 30% of dogs that enter the shelter system in Australia are euthanased and rehoming rates of 21% (Marston and Bennett, 2005) and 27.2% have been reported (Royal Society for the Prevention of Cruelty to Animals Australia National Statistics, 2010–2011). Worldwide, millions of dogs enter the welfare shelter system (Bollen and Horowitz, 2008) so there is a strong imperative for behavioural assessments to be valid. Deficiencies in assessment have the potential to be a significant welfare concern if dogs are wrongly euthanased. They also may expose members of the general public to harm, if aggressive dogs are inadvertently sold as pets.

The quality of a behavioural test (whether a test is a good measure, the right measure and a useful measure) is determined by three characteristics: reliability; validity; and feasibility (Martin and Bateson, 1993). It is critical, therefore, that behavioural assessment protocols used to determine the adoptability of shelter dogs are supported by empirical evidence to show that they meet accepted criteria regarding these characteristics (Taylor and Mills, 2006). At the very least, agreement between experienced raters conducting independent assessments of the same dog at the same time (inter-rater reliability) needs to be demonstrated, as does test–retest reliability, where the same dog is tested using an identical test on two or more occasions. In the case of shelter dogs, predictive validity is also of particular importance because, as identified by Duffy and Serpell (2012), behavioural tests are typically used to make generalisations about how dogs will behave in other environments, based on a limited sample of the dogs’ complete behavioural repertoire, observed within a short time period. This need for shelter tests to permit inferences about future dog behaviour makes thorough validation of such tests critical. In tests of predictive validity, rather than reassessing dogs in the original test location, one probes the experiences of new owners who have lived with the dog for a period of time. This enables questions to be asked about the dog's behaviour in a range of normal everyday situations (Duffy and Serpell, 2012) and provides an opportunity to explore the relationship the owner has formed with their new dog and whether this could have been predicted by the results of the original, in-shelter, assessment.

A review of shelter dog assessment protocols used in Australia (Mornement et al., 2010) revealed that shelters have made commendable attempts to ensure that only appropriate dogs are adopted. However, standardisation in assessment content and methodology, and empirical evidence to support the reliability, validity and feasibility of such protocols, is lacking. The aim in this study was to begin to address these issues by developing a scientifically informed and evaluated shelter dog assessment protocol, for use in Australia and elsewhere. Here we describe the initial development of the Behavioural Assessment for re-homing K9's (B.A.R.K.) and report on inter-rater reliability, test–retest reliability, predictive validity and feasibility of the instrument.

Section snippets

Focus group

Nine canine experts, including dog trainers (whose qualifications were at minimum, a Certificate III in Dog Behaviour & Training), a dog breeder, a veterinary behaviourist, an animal shelter manager, a shelter dog assessment officer, an animal welfare government representative and several research academics, attended a focus group. The aim of the focus group was to ascertain which behaviours were crucial for inclusion in a standardised behavioural assessment protocol for assessing adoption

Inter-rater reliability

Mean scores from both assessors for each behavioural attribute (anxiety, compliance, fear, friendliness and activity level) were obtained for each of the 12 subtests and overall. See Table 1 for Pearson's correlations of mean scores for the two assessors. The percentage agreement (PA) between the two raters is also presented.

As can be seen from Table 1, almost all correlations were statistically significant, but they ranged in size from weak to strong, based on Cohen's criteria (Pallant, 2007).

Discussion

The aims in this study were to develop a standardised behavioural assessment protocol for use with shelter dogs, and to evaluate aspects of the reliability, validity and feasibility of the instrument. The B.A.R.K. protocol was developed on the basis of previous research that investigated shelter dog assessment protocols currently used in Australia (Mornement et al., 2010) and a focus group with canine experts. This research showed that, although shelters routinely assess dogs for their adoptive

Conclusion

This study indicated that the B.A.R.K. protocol may be a useful tool in identifying critical behavioural traits, specifically fear and friendliness, in shelter dogs. However, the fact that the test–retest reliability and overall predictive validity were low is of concern, suggesting that the behaviours observed during a single behavioural assessment may fluctuate over time. Further research is therefore urgently required to investigate the test–retest reliability and predictive validity of

References (27)

Cited by (20)

  • Multi-Operator Qualitative Behavioural Assessment for dogs entering the shelter

    2019, Applied Animal Behaviour Science
    Citation Excerpt :

    An assessment of specific behavioural traits by attributing a score from 0 (absence of the behaviour, i.e. non-sociable) to 5 (extreme presentation of the behaviour, i.e. extremely sociable). The descriptors (sociability, calmness, fear, excitability, aggressiveness) were chosen based on animal temperament and personality studies, in particular in the literature describing behavioural traits of shelter dogs (Dowling-Guyer et al., 2011; Mornement et al., 2014; Svartberg and Forkman, 2002). For each session, the specific behavioural traits were evaluated in two different phases by the animal control officer (Animal Control Officer Form) and by the veterinary (Veterinary Officer Form; Table 1).

  • Consistency in behavior: Evaluation of behavior tests in laboratory beagles

    2017, Journal of Veterinary Behavior: Clinical Applications and Research
  • Test-retest reliability and predictive validity of a juvenile guide dog behavior test

    2016, Journal of Veterinary Behavior: Clinical Applications and Research
  • Evaluation of the predictive validity of the Behavioural Assessment for Re-homing K9's (B.A.R.K.) protocol and owner satisfaction with adopted dogs

    2015, Applied Animal Behaviour Science
    Citation Excerpt :

    The aims in this paper were to evaluate, more specifically, the predictive validity of the B.A.R.K. protocol by investigating, via a questionnaire, a range of behaviour exhibited by adopted dogs and how satisfied owners were with their new dogs and the adoption process. Preliminary evaluation of a general measure of the predictive validity of the instrument had previously shown that overall mean scores for fear and friendliness obtained from the B.A.R.K. protocol correlated with ratings given by new owners, for the same behaviours, post adoption (Mornement et al., 2014). Upon investigation of the ability of the B.A.R.K. protocol to predict more specific behaviours post adoption, the results were somewhat disappointing.

  • Applied personality assessment in domestic dogs: Limitations and caveats

    2015, Applied Animal Behaviour Science
    Citation Excerpt :

    Tests which only assess gross, active behaviours and are aimed at assessing personality or predicting the potential for future aggressive or fearful reactions, rely on the incorrect assumption that individuals displaying passive behaviour are not experiencing considerable stress or arousal and will suffer from poor predictability due to these inherent design flaws. Some support for this idea is shown by the degree to which measures of passive behaviour increase the accuracy of tests; battery tests scored primarily on the presence/absence of active responses show lower levels of concurrent validity or predictability validity for behaviours related to aggression and fear (Bennett et al., 2012; Mornement et al., 2014), than those that include measures of subtle, passive behaviours that are thought to be indicative of stress (Barnard et al., 2012; Kuhne et al., 2014). Given the methodological limitations inherent to assessments carried out in shelters and welfare facilities, and to a degree in other areas in which canine personality tests are used, it appears that behaviour-only tests will always suffer from validity and reliability issues due to the trade-off between the high level of detail required to examine all relevant behaviours and making sure tests are both transferable and feasible.

View all citing articles on Scopus
View full text