Our researchers argue that behavior evaluations for shelter dogs can never be more reliable than flipping a coin, even under unrealistically optimistic, hypothetical conditions.

They conducted a review of the literature on dog bites and owner surrenders and an analysis of the data. Their goal was to determine whether even a very good evaluation was statistically likely to yield results that might be helpful in making decisions about individual dogs. The result: it was not.

The researchers advocate against formal dog evaluations in shelters and instead advocate for focusing on behavioral histories, incident verification, and prioritizing positive activities where a dog’s social skills can be observed and practiced.