Validation of the Abbreviated C-BARQ Used by the Dog Aging Project to Describe Canine Behavior
INTRO:
In human and veterinary medicine, the study of behavior is a perpetually evolving field. While other fields of research work with more tangible data, behaviorists are faced with a daunting task – to observe and interpret complex behaviors, without specific lab work or biomarkers to assist them (1). In humans, this is often completed via in-depth questionnaires, where specific patterns or tendencies may help to identify or describe an abnormal behavior. However, in veterinary medicine, there is the distinct disadvantage that the dog displaying the abnormal behavior is not capable of self-reporting.
The behavioral profiles of animals are often determined via observation. A complicating factor of this process is that the mere presence of a researcher can be sufficient to significantly modify an animal’s behavior, particularly if the animal is uncomfortable or aggressive towards strangers (2). This principle also applies to the environment in which the observation occurs, as a veterinary hospital or research laboratory setting itself can result in abnormal behaviors (2). Because of these intricacies, a need arose to develop methods for obtaining information that could be performed by someone with intimate knowledge of the pet who could observe the pet extensively without causing any undue stress.
In response to this need, canine behavioral questionnaires have been developed to facilitate a behavioral diagnosis based on owner reports. Multiple behavior surveys exist, including the Monash Canine Personality Questionnaire (MCPQ), the Dog Personality Questionnaire, and the Canine Behavioral Assessment and Research Questionnaire (C-BARQ) (3-5). The original C-BARQ is a 100-item questionnaire designed to investigate dog behavior along 14 different behavioral domains (6, 7). The C-BARQ is perhaps the most commonly used canine behavior assessment tool throughout the world, and has been studied and used in Japan, Latin America, Iran, and the United States (2, 8-10). The C-BARQ has been utilized to determine how canine behavior is influenced by diverse variables such as early life experiences; manner of acquisition, e.g., purchase from breeders or pet stores; and the breed of dog (5, 11, 12). The C-BARQ has also been able to provide valuable information regarding a dog’s likelihood of future success as a guide dog, and has been used in shelter practice to assess the adoption potential of recently-relinquished dogs (13, 14).
Due to the length of the original C-BARQ, concerns regarding ease of use and owner compliance prompted the development of shortened, or abbreviated, versions of this tool. One such abbreviated C-BARQ has been used to describe the behavioral results of puppy farming (7). While this shortened version reportedly provides reliably similar behavioral assessments when compared to the original, its results have never been directly compared to those obtained from the full-length version. When Duffy, et al. created their own version of an abbreviated C-BARQ (C-BARQ(S)), they assessed its validity by retroactively comparing the newly generated results to those obtained from the standard questionnaire (13). In their study investigating the effects of puppy farming on dog behavior, Wauthier, et al. used the same shortened questionnaire without further validation (7). Thus, the validity of an abbreviated CBARQ has never been investigated via the administration of both CBARQ versions to the same human-dog pair.
METHODS:
The ‘abbreviated’ C-BARQ (a-CBARQ) comprises the behavioral section of the Health and Life Experience Survey (HLES), an extensive survey completed by all members of the Dog Aging Project (DAP) pack. At this time, there are currently over 300 study participants who have completed the HLES, including the a-CBARQ. Due to the fact that these participants have already been made aware that they will be contacted regarding future studies, and because they have already shown a willingness to complete online surveys, they represent an ideal group to complete the original CBARQ for comparison of results between the two scales. All current DAP Pack members will receive an invitation to create a C-BARQ account within the original, online C-BARQ platform, using a shared code unique to the DAP, and will be asked to complete the full-length C-BARQ questionnaire. Their responses upon completion of this version of the survey will be compared to those obtained from the a-CBARQ already completed on our DAP platform as part of the HLES survey. This number of participants is similar to other studies that use other canine behavioral questionnaires, such as the DPQ (217 initial surveys completed in Turcsan, et al. (4)) or the MCPQ (588 individuals in Ley, et al. (3)). The other mini-CBARQ study (Wauthier, et al.) involved over 2000 participants, but none of the participants also completed the full-length questionnaire for comparison of results (7).
ANALYSIS:
A comparison of the original with the abbreviated CBARQ will be made between all DAP pack members who complete both forms of the instrument (expected to reach over 800 members). Descriptive statistics will include means/SD and counts/% for dog characteristics including age, breed, sex, and geographic residence. We will conduct an exploratory Principal Components Analysis (PCA) of scores from the original CBARQ to better understand collinearity and factor loadings within domains of behavior. Agreement between the two versions of the questionnaire will be based on analysis of scores (each question rated from 0 to 4) within the 14 domains designated within the CBARQ questionnaire: stranger-directed aggression, owner-directed aggression, dog-directed aggression, dog rivalry, stranger-directed fear, nonsocial fear, dog-directed fear, separation-related behavior, attachment and attention-seeking, trainability, chasing, excitability, touch sensitivity, and energy level. Standardized scores (mean of 0 and SD of 1) will be calculated for each domain in both instruments and rounded to the nearest whole number. A weighted kappa for ordinal data will be used to determine level of agreement between the two scores (15). We will interpret agreement between the two CBARQ instruments using conventional criteria: ≤ 0 as indicating no agreement, 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41– 0.60 as moderate, 0.61–0.80 as substantial, and 0.81–1.00 as almost perfect agreement (16).
REFERENCES:
1. Borchelt, P.L. and V.L. Voith, Diagnosis and Treatment of Separation-Related Behavior Problems in Dogs. Veterinary Clinics of North America: Small Animal Practice, 1982. 12(4): p. 625-635.
2. Wiener, P. and M.J. Haskell, Use of questionnaire-based data to assess dog personality. Journal of Veterinary Behavior, 2016. 16: p. 81-85.
3. Ley, J.M., P.C. Bennett, and G.J. Coleman, A refinement and validation of the Monash Canine Personality Questionnaire (MCPQ). Applied Animal Behaviour Science, 2009. 116(2-4): p. 220-227.
4. Turcsan, B., et al., Personality traits in companion dogs-Results from the VIDOPET. PLoS One, 2018. 13(4): p. e0195448.
5. van den Berg, S.M., et al., Evaluation of the C-BARQ as a measure of stranger-directed aggression in three common dog breeds. Applied Animal Behaviour Science, 2010. 124(3-4): p. 136-141.
6. Hsu, Y. and J.A. Serpell, Development and validation of a questionnaire for measuring behavior and temperament traits in pet dogs. J Am Vet Med Assoc, 2003. 223(9): p. 1293-300.
7. Wauthier, L.M. and J.M. Williams, Using the Mini C-BARQ to Investigate the Effects of Puppy Farming on Dog Behaviour. Applied Animal Behaviour Science, 2018.
8. Tamimi, N., et al., Assessment of the C-BARQ for evaluating dog behavior in Iran. Journal of Veterinary Behavior, 2015. 10(1): p. 36-40.
9. González-Ramírez, M.T., L. Quezada-Berumen, and R. Landero-Hernández, Assessment of canine behaviors using C-BARQ in a sample from Northern Mexico. Journal of Veterinary Behavior, 2017. 20: p. 52-58.
10. Nagasawa, M., et al., Assessment of the factorial structures of the C-BARQ in Japan. J Vet Med Sci, 2011. 73(7): p. 869-75.
11. Foyer, P., et al., Behaviour and experiences of dogs during the first year of life predict the outcome in a later temperament test. Applied Animal Behaviour Science, 2014. 155: p. 93-100.
12. McMillan, F., et al., Differences in behavioral characteristics between dogs obtained as puppies from pet stores and those obtained from noncommercial breeders. JAVMA, 2013. 242.
13. Duffy, D.L., K.A. Kruger, and J.A. Serpell, Evaluation of a behavioral assessment tool for dogs relinquished to shelters. Prev Vet Med, 2014. 117(3-4): p. 601-9.
14. Serpell, J.A. and Y. Hsu, Development and validation of a novel method for evaluating behavior and temperament in guide dogs. Applied Animal Behaviour Science, 2001. 72(4): p. 347-364.
15. Ranganathan P, Pramesh CS, Aggarwal R. Common pitfalls in statistical analysis: Measures of agreement. Perspect Clin Res. 2017;8(4):187‐191
16. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb). 2012;22(3):276‐282.