Evaluation of behaviour testing for human directed aggression in dogs

Abstract

Behaviour test batteries are used to identify aggressive dogs. The Dutch Socially Acceptable Behaviour (SAB)-test has been used since 2001 to select against unwanted aggression and fear in specific dog breeds, though much is unknown yet regarding its reliability, validity and feasibility. In this paper the SAB-test is evaluated and the results describe its usefulness for identifying aggression towards humans in dogs. Behaviour test outcomes on 345 dogs (479 records on tests performed indoor and outdoor) scored by the judges of the Dutch Kennel Club were compared to owner reported assessments of their dogs’ behaviour prior to testing. Dogs were labelled aggressive when having bitten a human at least once according to the owner and were diagnosed by the judges as such when attacking at least once during the SAB-test. Aggressive dogs showed significantly more attacks than non-aggressive controls and a Principal Components Analysis of detailed observations on 76 dogs grouped aggressive behaviours like growl, bare teeth, snap, lunge and bite in one dimension, confirming the test's validity. Analysis of 479 test records revealed a sensitivity, specificity and accuracy of 0.33, 0.81 and 0.64, respectively. The specificity could be increased to 0.93 by raising the threshold for positive test results to at least 2 attacks during the test. The low sensitivity is explained by the decision to classify dogs as aggressive on the basis of one bite incident only and by a relative weak capacity of the test to detect specific forms of aggression. The reliability of the test was investigated by comparing test outcomes on 133 dogs when tested indoor and outdoor. The accuracy decreased from 0.67 to 0.62, but overall the indoor test outcomes were similar to those found outdoors. Scores for aggression and anxiety were significantly higher when dogs were tested for the first time in the morning than the second time in the afternoon, suggesting desensitization. Salivary cortisol concentrations in 20 dogs were not different in samples taken before and after the test, which oppose severe levels of stress. The SAB-test allows one to evaluate aggression in dogs, but present findings indicate that probably those that behave aggressively in the absence of fear remain undetected. It is recommended to increase the test's usefulness by refining or including new test components that target different forms of aggression. Formulating a risk assessment based on detailed information on a dog's behaviour during testing instead of simply producing a pass–fail judgement will facilitate a purpose specific use of the SAB-test.