Shelter Dogs And The Approach Test
Since dogs are domesticated animals, their relationships with humans have a large impact on their overall welfare. A dog’s approach behavior (how the dog reacts towards a stranger human who is approaching in a non-threatening way) may therefore serve as an indicator of this relationship and of the dog’s welfare. Approach behavior is especially important for shelter dogs, whose level of willingness to meet strangers affects their chances of finding a home.
This study, published by Elsevier in the journal of Applied Animal Behavior Science, looked to assess how well an approach test actually evaluates dog welfare in terms of giving results that are both stable and valid. As part of the study, experimenters visited 29 different shelters to test a total of 520 dogs. Dogs who reacted to a non-threatening stranger’s approach with non-aggressive curiosity were categorized as “contact possible” whereas dogs who reacted by hiding, ignoring the stranger, barking or growling continuously, or attacking were categorized as “no contact possible.” In addition, staff at each shelter filled out a survey answering questions related to their attitude towards their work and towards dogs. The authors wanted to see how the shelter staffs’ attitudes impacted the approach behavior of the dogs at their shelter.
Of the 520 dogs tested, 69% were categorized as “contact possible” and the rest (31%) as “no contact possible.” To test for stability of the approach test results between experimenters, a total of 158 dogs from 7 different shelters were tested twice by two different experimenters in the same visit. The authors found that 95% of these dogs were assigned to the same category by both experimenters. The authors additionally tested for stability of the approach test results over time: 9 shelters were visited twice, with 50-76 days between visits. On an individual level, 87% of dogs were assigned to the same category on both visits. The authors analyzed results between visits and concluded that retest reliability was acceptable and that there was no significant difference between results from the two visits.
Validity of the test results was determined by combining information from the approach tests with results from shelter staff surveys. The authors found that over ¾ of survey respondents agreed with positive adjectives (such as “playful” or “cuddly”) to describe dogs; these respondents were more likely to have reported being more comfortable interacting with dogs and to have agreed more with the use of gentle and predictable dog handling practices. In contrast, respondents who agreed with negative adjectives (such as “noisy” or smelly”) to describe dogs made up less than ¼ of respondents and were more likely to agree with coercive dog handling practices. Additionally, ½ of respondents found their job stressful but nevertheless ¾ found it pleasant; the ½ who found their job stressful were more likely to agree with coercive handling practices.
Upon looking at survey responses alongside approach test results, the authors found that in all but one subsample of survey respondents, there was a correlation between higher proportions of dogs approaching experimenters and shelter staff reporting lower levels of comfort interacting with dogs. In the subsample of staff who worked over 80% of their time with dogs only, the authors instead found a correlation between higher proportions of dogs approaching experimenters and staff reporting more positive attitudes towards dogs as well as higher levels of agreement with gentle and predictable handling practices. The authors hypothesize that in the first case the dogs who approach strangers do so because they do not receive enough human interaction from shelter staff, whereas in the second case the dogs who approach strangers do so because they have been socialized to be friendly with humans by positively-minded shelter staff. These conflicting factors impair the validity of using the approach test alone to determine dogs’ welfare since higher approach behavior indicates lower welfare in the first case but higher welfare in the second.
The conclusions of the authors are as follows: Having caretakers with a positive attitude towards dogs appears to be important for the dogs’ welfare. This is especially true since people with negative attitudes towards dogs tend to agree more with coercive handling methods, which previous studies have found lead to more aggressive and fearful dogs. As for the use of approach tests to assess welfare, the authors conclude that the approach test gives stable results. However, validity is a problem since high approach behavior can be an indicator of either low or high welfare. The authors therefore suggest a combination of approach tests with other measures to assess dog welfare.
[Contributed by Mona Zahir]