Shelter C-BARQ Evaluation Research

Assessing the behavior of dogs while in the shelter has long been a complicated task. Research has shown mixed results about the validity and reliability of behavior evaluations based on observing a dog’s behavior across various “test items” in a single setting. These behavior evaluations provide only partial information about a dog’s behavior in a particular context (e.g., in shelter) and with nonfamiliar people handling the dog. It is best practice to gather information from multiple people across various contexts (e.g., in kennel, in foster care, on a walk) to identify an animal’s typical behavior. All of this takes a significant amount of time and resources and may be of limited use in predicting a dog’s behavior in a home environment. Thus, efficiently getting accurate behavioral information, especially about the presence or severity of problem behavior in individual animals, is still challenging

Hsu & Serpell developed the Canine Behavioral Assessment and Research Questionnaire (C-BARQ) to identify the presence and severity of problem behaviors by surveying dog owners about their dog’s behavior. Duffy et. al. (2014) evaluated a shortened version of the C-BARQ to determine its value as an intake survey instrument by asking relinquishing owners and adopters of the same dog to complete the survey. While promising, this application of the C-BARQ still relied on behavior observations of the dog in a home setting, not in the shelter. Gilchrist, et. al. (2025) designed a research project to determine if the C-BARQ could be used to accurately identify problem behaviors for dogs living in the shelter by asking staff and volunteers to complete a version of the C-BARQ. Could a short survey give us the necessary information to make wise pathway and outcome decisions, and do so efficiently?

What did they do?

Researchers asked staff from 11 different shelters who best knew each dog in the study (n=445) to complete a subset of the CBARQ while the dog was living in the shelter. The questions included in this C-BARQ subset (37) could be asked about dogs living in a shelter or home. They split the data into two groups (i.e., the exploratory and the confirmatory groups) before analysis. This allowed them to analyze the exploratory group to see which questions were grouped reliably in different behavior categories. Once they knew which questions contributed to which behavior categories, they ran a different analysis on the confirmatory group. This approach allowed them to identify, evaluate, and validate the use of a subset of C-BARQ questions for sheltered dogs to assess behavior while they live in the shelter.

Who was included in the study? (aka, “The Sample”)

When evaluating the applicability of any research paper, it is essential to know who the researchers asked about their research question. In other words, who contributed the information to answer the research question? This is referred to as the “sample”. The more closely the research sample is to your own population, the more likely the results can be applied in your shelter. In this study, we need to look at two different “samples” – the people filling out the C-BARQ and the dogs about whom they filled out the survey. Previously, the C-BARQ relied on people who lived with the dog to complete the survey based on knowledge they gained about the dog while sharing the same space with the dog. For this project, staff who best knew the dog completed the C-BARQ based on their interactions with the dog while the dog lived in the shelter. They did not share the same housing space with the dog. There was no minimum requirement for how long the staff member knew the dog or the number of interactions with the dog about whom they filled out the survey. We do not know the demographics of the people who took the survey. Therefore, we don’t know how well they match your shelter staff characteristics. However, they are shelter staff, so we can reasonably assume that they may share some similarities with staff at your shelter. Most dogs in this research project were “mature adult male dogs of medium to large size” who did not display extreme or dangerous behavior. Thus, we know that the results of this study are limited to dogs of a specific sex (i.e., male) and size (i.e., medium to large). We need to use caution when applying the results of this study to other dogs, such as female dogs, small dogs, puppies, and dogs displaying extreme or dangerous behavior. More research to confirm that the C-BARQ is equally accurate for those dogs in those categories will increase confidence in its applicability across a variety of dogs.

Results

Analysis identified twenty-four questions that were grouped into five behavior categories: 

• Fear (seven questions): most precise for identifying dogs with high levels of fear
• Arousal (six items): most informative for dogs with average arousal levels
• Human Excitability (four questions): most precise for dogs that show excitement slightly above or below average
• Dog Aggression (four questions): most precise for dogs that show average to high levels of aggression
• Human Aggression (three questions): effectively identifies dogs with high levels of human-directed aggression 

Based on this analysis, a Shelter C-BARQ was developed to provide insight into a shelter-housed dog’s behavior. Each behavior category has its own scale. Thus, you can add up the scores within each category, but you do not add the category scores for an overall behavior score. Shelter staff completing the shelter C-BARQ choose a dog’s recent tendency to display behavior from each category on a scale of 1 (None) to 5 (Extreme) in specific contexts. For the exact questions in each behavior category in the Shelter C-BARQ, see the Appendix of the original research paper.

The Takeaway

Assessing sheltered dogs’ behavior is still an important responsibility for shelters. The challenge of combining efficiency with accuracy has resulted in many shelters designing their own tools and methods to gather behavioral information to inform inshelter behavioral support, pathway planning and outcome decisions, and placements. Creating tools and methods that can be performed quickly and efficiently may make it more feasible for shelter staff to use them. However, doing so sacrifices the validity and reliability of the information gathered. It is best to use tools and methods of using those tools that are already validated by research and shown to provide reliable, accurate information. When you change a tool or its use to fit your needs, it is no longer a validated process of gathering behavioral information.

While scientifically validated, the shelter C-BARQ still needs additional research to increase applicability across shelter staff and diverse dog populations. For example, this study looked at one staff member’s rating of a dog. Researchers did not test how well this tool works when multiple people fill out the survey about the same dog. They did not ask whether the tool's accuracy improves with more interactions over a longer period with the person completing the survey. And some human factors could contribute to how they choose their ratings, such as age, gender, experience with dogs, primary role in the shelter (e.g., medical, behavior, daily care, operations), etc. 

If you are already doing a formal, scientifically validated behavior assessment, this tool may not yield additional information from your current behavior assessment process. For shelters not currently using a scientifically validated behavior assessment tool, the Shelter C-BARQ is an efficient, scientifically validated option for identifying shelter-housed dogs who may benefit from behavioral intervention or require pre-adoption counseling. The Shelter C-BARQ, or any tool, should be combined with ongoing behavior observation and tracking throughout the dog’s shelter stay to ensure appropriate behavior support is provided from intake to outcome.