Summary: An overview of available research on some of the temperament tests currently used as first-line intake evaluations for dogs in a shelter, why these tests are used at all, and some possible alternatives.

Animal shelters and rescues have two essential responsibilities relating to the dogs that are placed in their care: providing for the care and welfare of these dogs, and placing them in suitable homes.¹ These challenging missions are complicated by the need to determine whether the dogs have behavioral traits that stand in the way of successful adoption or present a danger to the shelter staff. Unfortunately, there are few effective tools to aid in assessing the dogs’ temperament and behavioral traits once they’re in a shelter.

No matter how well a shelter is designed, operated and maintained, it is still a foreign, frightening, and stressful place for the dogs that are housed in it. The dogs have been removed from whatever environment they considered home and are in an unfamiliar place, with strange noises, smells, and people, and are surrounded by other highly stressed animals. Upon arrival, they are given a physical examination and are often kept in quarantine with little human contact, then are moved to kennels where strangers come and go. And, with the dogs being handled under these difficult circumstances, the shelter staff is expected to determine whether they can be safely housed and cared for, and can be adopted into people’s homes. Shelters commonly use temperament, behavior, and personality tests as part of this process.

There are numerous problems associated with these tests. For starters, there is no agreement as to what the tests are intended to measure or what their intended purpose may be. Tests are often intended to assess dogs’ suitability for specific working roles, such as law enforcement or service dogs, rather than to determine whether a dog has behavioral problems.² An additional problem with such tests is that they are generally meant to be administered under conditions that are similar to the ones that the dogs would be working, rather than in a shelter environment. In fact, behavioral evaluations performed in shelters are done under conditions, and performed by people, that are completely unfamiliar to the dog being tested.³ Lastly, there is very little agreement as to the terminology and methods used in administering and evaluating the tests,⁴ making it very difficult to validate their results and determine their effectiveness in predicting dogs’ behavior.⁵

In-shelter testing is particularly problematic in assessing what behaviors, or behavioral problems, a dog will have in an adoptive home. While in the shelter, the dogs are reacting to unfamiliar and uncomfortable conditions, which are simply not present anywhere else.³ Research has shown that testing dogs’ reactions to stimuli under the adverse conditions in a shelter leads to completely inaccurate results.⁶ In fact, one study estimated the false positive error rate of in-shelter testing to be 68%.⁷ Upon reviewing the data associated with in-shelter temperament testing, the ASPCA has determined that they tend to result in false positive indicators of aggressive behavior, and recommends that they be considered valid only if corroborated by other sources.⁸

However, battery testing in shelters can be a useful tool in developing safety protocols for shelter staff who will be handling particular dogs and in caring for them while in the shelter. We can learn how the dogs tend to react to the novel stimuli that they will encounter in shelters,⁹ which is helpful in developing plans for their welfare and safe handling. We’ll come back to this later.

The tests used to assess animals’ temperament or personality generally consist of subjecting them to various stimuli and observing their reactions. However, the persons who are administering these tests are handicapped by the animals’ inability to provide the context of their reactions. Since the animals, in this case shelter dogs, are unable to clearly communicate their emotional state or their understanding of the test stimuli, the evaluators are limited in their ability to interpret the dogs’ reactions. Also, battery tests often fail to note or record various coping behaviors that dogs might use in response to test situations.⁵ For example, a dog may have resorted to “aggressive” behavior because all avoidance or distancing options were not available to them during the test. A 2020 study of in-shelter behavioral testing as a means of predicting post-adoption behavior had mixed results. Observed friendly and sociable or anxious and fearful behaviors were consistent in both environments; however, aggression, guarding, or separation anxiety were not reliably predicted.¹⁰

Testing a dog for food guarding, using a fake human hand. (The Science Dog (2013)

Studies have shown that certain behaviors that are considered to be “aggressive” are often associated with the fact that the dogs are being housed in shelters, and do not necessarily manifest in home environments. One such behavior, food guarding or guarding behavior in general, is often considered to be a high impediment to adoption. Dogs exhibiting this behavior are often placed on restricted adoption or are considered dangerous and are euthanized. However, research has found that there is no correlation between guarding in the shelter and in adoptive homes, and that the majority of adoptive owners did not consider the guarding behavior to be a serious problem.¹¹ In fact, researchers have found that removing any tests for guarding behavior did not result in injuries to shelter staff, nor did it have any relationship to the number of dogs that were returned to the shelter.¹² Another behavioral test that is often performed on shelter dogs is to evaluate their reaction to dummies that represent other dogs or small children. These tests are intended to determine whether the dogs are prone to aggression responses in the presence of other dogs or are suitable for placement in homes with children. The obvious limitation of these tests is that “these devices do not smell, move or interact like real children or conspecifics and tested dogs may perceive them as social stimuli only at a distance and for the very few first seconds of interaction.”¹³ Essentially, observers are expected to evaluate the dogs’ social interaction with inanimate objects that provide no behavioral cues. As mentioned earlier, these tests may be useful in evaluating a dog’s reaction to a novel situation or stimulus, but do not appear to provide any information about the dog’s reaction to unfamiliar humans or other dogs.

In the absence of reliable means of testing dogs’ temperament in an unfamiliar and stressful shelter environment, it has been suggested that better results could be obtained by obtaining a detailed history of the dog upon arrival at the shelter.⁷ Detailed information on dogs’ behavioral history is thought to be the best indicator of their post-adoption behavior and temperament, and the best means of successfully matching them with an adoptive home. The issue then becomes how to obtain an accurate history. Shelters commonly interview dogs’ owners during the intake process; however, this can result in the shelter receiving misleading information. When surrendering their dog, owners can be less than honest about serious behavior problems in the belief that this will increase the dog’s chances of adoption.¹⁴ We must also keep in mind that surrendering a pet is an extremely emotional process for owners, during which they can feel judged or defensive about their actions and reasons for relinquishment, thus coloring their interactions with shelter staff.

It has been suggested that the better means of obtaining an accurate behavior history from people who are surrendering their dogs is by means of a questionnaire. This allows the owners to provide information about the dogs’ actions and characteristics without the emotional burden of in-person discussions. While the questionnaires may not predict the behavior of dogs in the unfamiliar and stressful environment of an animal shelter, they may provide very useful insight as to their behavior in an adoptive home. Although some research has shown that questionnaires are not consistent in predicting future behavior,¹⁵ two have been assessed as having a high degree of reliability. They are a modified version of the C-BARQ personality assessment and the Match-Up II test developed by the Animal Rescue League of Boston.

In many cases, dogs are placed in shelters with no history and with no information about their past behavior. Dogs are often transferred between shelters to reduce crowding and to increase their chance of adoption, many times with incomplete information. Shelters also receive stray or abandoned dogs from local authorities or good Samaritans who find them running loose. In these cases, shelters have no way of knowing what the dogs’ behaviors were in their previous environments and no real way to predict how they will perform in adoptive homes. For these dogs, the emphasis must be to assess their behaviors in the shelter environment, to aid in their well-being, and to protect their handlers from harm. Menchetti et al suggest that this can be accomplished by having multiple qualified persons perform a brief standardized evaluation during various phases of intake and care.¹⁶ The study suggests that the animal control officer and shelter veterinarian perform qualitative evaluations of the dogs’ actions using the same form and criteria, and that this same evaluation be periodically performed by knowledgeable members of the shelter’s animal care staff. It is suggested that this process provides not only a consistent evaluation by multiple parties but allows the shelter to identify and document any changes.

Several qualitative evaluations by qualified individuals over time may be the best way to assess a dog’s behavior and track changes.

Summing it up: There is a clear need for shelters to have a means of determining the type and severity of any behavioral problems that their dogs are displaying, to help place the dogs in adoptive homes, aid the shelter staff in caring for the dogs, and to protect the shelter staff from reactive or aggressive animals. However, there is a lack of effective tools available to them. Battery tests are of very questionable value in predicting the behavior of dogs after they are placed in homes, and certain components of these tests – notably those related to resource guarding – have been shown to be no better than random chance. The available research indicates that the best means of predicting post-adoption behavior is to ask dog owners to fill out a detailed questionnaire when relinquishing their pet, and the best means of predicting future behaviors while a dog is in the shelter is to have multiple operators perform qualitative assessments during intake and periodically afterward.

References

1. ASPCA (nd). Position Statement on Shelter Dog Behavior Assessments. Last accessed 10/7/2023.
2. Taylor, K.T. & Mills, D.S. (2006). The development and assessment of temperament tests for adult companion dogs. Journal of Veterinary Behavior 1:3, 94-108
3. Rayment, D.J., De Groef, B., Peters, R.A., & Marston, L. C. (2015). Applied personality assessment in domestic dogs: limitations and caveats. Journal of Applied Animal Behaviour Science 163, 1-18.
4. Brady, K., Cracknell, N., Zulch, H., & Mills, D. S. (2018). A systematic review of the reliability and validity of behavioural tests used to assess behavioural characteristics important in working dogs. Frontiers in Veterinary Science 5:103.
5. Diederich, C. & Giffroy, J. (2006). Behavioural testing in dogs: A review of methodology in search of standardization. Journal of Applied Animal Behaviour Science 97, 51-72.
6. Patronek, G. & Bradley, J. (2016). No better than flipping a coin: Reconsidering canine behavior evaluations in animal shelters. Journal of Veterinary Behavior 15, 66-77.
7. Patronek, G. J., Bradley, J., & Arps, E. (2019). What is the evidence for reliability and validity of behavior evaluations for shelter dogs? A prequel to “No better than flipping a coin”. Journal of Veterinary Behavior 31. 43-58.
8. Haverbeke, A., Pluijmakers, J., & Diederich, C. (2014). Behavioral evaluations of shelter dogs: literature review, perspectives and follow-up with European member states’ legislation with emphasis on the Belgian situation. Journal of Veterinary Behavior 10, 5-11.
9. Clay, M., Materson, M.B.A., Bennett, P., Perry, G. and Phillips, C.C J. (2020). Do behaviour assessments in a shelter predict the behaviour of dogs post-adoption? Animals 10:7, 1225.
10. Marder, A.R., Shabelansku, A., Patronek, G.J., Dowling-Guyer, S. & D’Arpino S.S. (2013). food-related aggression in shelter dogs: A comparison of behavior identified by a behavior evaluation in the shelter and owner reports after adoption. Journal of Applied Animal Behaviour Science 148:1-2, 150-156.
11. Mohan-Gibbons, H., Dolan, E.D., Reid, P., Slater, M.R., Mulligan, H. & Weiss, E. (2017). The impact of excluding food guarding from a standardized behavioral canine assessment in animal shelters. Animals 8:2, 27.
12. Barnard, S., Siracusa, C., Reisner, I., Valsecchi, P & Serpell, J. (2012). Validity of model devices used to assess canine temperament in behavioral tests. Applied Animal Behaviour Science 138:1-2, 79-87.
13. Sergurson, S.A., Serpell, J.A. and Hart, B.L. (2005) Evaluation of a behavioral assessment questionnaire for use in the characterization of behavioral problems of dogs relinquished to animal shelters. Journal of the American Veterinary Association 227:11, 1755–1761.
14. Duffy, D.L., Kruger, K.A. & Serpell, J.A. (2014). Evaluation of a behavioral assessment tool for dogs relinquished to shelters. Preventive Veterinary Medicine 117 (2014). 601-609.
15. Dowling-Guyer, S., Marder, A. & D’Arpino, S. (2010). Behavioral traits detected in shelter dogs by a behavior evaluation. Journal of Applied Animal Behaviour Science 130:3-4, 107-114.
16. Menchetti, L., Righi, C., Geulfi, G., Enas, C., Moscati, L., Mancini, S. & Diverio, S. (2019). Multi-operator qualitative behavioural assessment for dogs entering the shelter. Journal of Applied Animal Behaviour Science. 213(2019). 107-116.

John Reilly, MS, CBCC-KA, is a volunteer dog trainer and behavior consultant for the Potter League for Animals in Middletown, Rhode Island, and runs the training program for volunteers who assist with the shelter’s dogs.  He is also the owner and primary blogger at www.animalnerd.com