A commonly used "food bowl" test â€” in which the tester removes a food bowl from a dog to test whether the animal shows extreme or dangerous levels of "resource guarding" behavior â€” has been shown, according to studies cited in theTimesarticle, to be an unreliable predictor of how a dog is likely to behave in the home.
The implications of the finding are significant, since how a dog fares on the "temp test" can make the difference between getting placed with a family and getting put to sleep. Unreliable results can result in a harmless dog being euthanized or a family being put at risk.
However important, the new findings on temp test reliability are hardly unexpected. It's just not that surprising that how a dog acts under test conditions â€” in the stress of a shelter, having its food swiped by a plastic hand â€” is not a reliable guide to how it would behave in the safety of a home environment. When it comes to behavior â€” human as well as animal â€” context matters.
But it isn't that the test is meaningless. As Sara Scott, a certified professional dog trainer in the San Francisco area's East Bay (whom I have hired to work with my own dog) explained to me by telephone, probes such as the food bowl test do give you a snapshot of how the dog is doing in the shelter.
"But behavior is fluid and not static," Scott said. "A dog's behavior changes from month to month as its situation changes. And how a dog behaves in the home is going to have as much to do with how the owners behave, with the kind of training and management they offer the dog. A lot of problematic behavior comes from the fact that a dog doesn't feel safe, and isn't having their physical and mental needs met."
The problem is not so much the test as a too single-minded focus on the test alone. As Scott explained, an experienced shelter worker can usually tell right off the bat that a dog has resource guarding issues.
As Scott put it: "It's probably clear in the dog's bearing right from the start, in the way he stiffens at your approach, or starts gulping down food in a hurry, or growls."
The test comes in handy, though, when a dog has learned to conceal its anxiety and passes over growling and goes right to the biting stage, she said. And the plastic hand is useful in a case like that because it keeps the shelter worker safe.
But for dog placements to succeed, you can't rely on tests alone. You need equal attention on finding adopters who are appropriate to the animal â€” and then provide them the education and resources they need to protect the dog.
The thing about shelters, Scott continued, is that staff are overworked and may not have the time or resources to support either the proper assessment of the animal, or the proper selection and education of adoptive families. Indeed, the problem may be worse: Shelter workers are sometimes underpaid; there's lots of staff turnover; experienced dog people may leave to find better-paying work elsewhere. So the result is that you are more likely to have shelter workers lacking the training to make either effective use of the temperaments tests or to ensure that the right adopters are found for a dog.
All with the consequence that life-and-death decisions are made on the basis of tests that can hardly be expected to bear that weight all by themselves. Which brings us to the all-too-predictable finding that test results aren't necessarily reliable predictors of adoption success.
The real issue is the need to give handlers, shelters â€” and the whole process â€” the resources they need so that, finally, dogs, and people, are supported and kept safe.
And crucially, as Scott emphasized: "We have too many animals â€” and it's a human-created problem. It's our responsibility to take care of it. Not all the dogs are going to be successful in the home, so it becomes a numbers game. You've got to pick the dogs that will be successful, so you can save more lives."