HABIC2 summary

Slide1: 

Good Dogs and Bad: How to Evaluate the Quality of Canine Companionship James A. Serpell, PhD Center for the Interaction of Animals and Society, School of Veterinary Medicine, University of Pennsylvania

Pet ownership is like ‘marriage’ – not all human-animal bonds are made in heaven.: 

Pet ownership is like ‘marriage’ – not all human-animal bonds are made in heaven.

Slide3: 

Relationships with pets are like relationships with people People’s attachments for their dogs are strongly influenced by the animal’s behavior. (Serpell, J.A. 1996. Appl. Anim. Behav. Sci., 47: 49-60). Pet owners whose pets are behaviorally ‘compatible’ report better mental and physical health than those with less compatible pets. (Budge, R.C. et al. 1998. Society & Animals, 6(3), 219-234) Positive relationship provisions provided by pets are comparable to, and sometimes better than, those provided by other people. (Bonas et al. 2000. In: Companion Animals & Us, eds. A.L. Podberscek et al., pp. 209-236, Cambridge: CUP). Negative relationship provisions derived from pets are less extreme than those derived from other people. (Bonas et al. 2000. In: Companion Animals & Us, eds. A.L. Podberscek et al., pp. 209-236, Cambridge: CUP).

Slide4: 

Owners of dogs with behavioral problems or deficits probably derive less social support, and fewer health benefits from their animal’s companionship. Owners will be more likely to disown dogs with behavioral problems or deficits (behavior problems account for ± 25% of all shelter relinquishments). Owners of problem dogs will be more likely to resort to inappropriate treatment/punishment. This implies that…….

Slide5: 

Therefore……….. We need a better understanding of the origins (genetic, developmental) of behavioral problems in the canine population. But before we can achieve this……. We need to develop standardized and widely accepted methods for describing and evaluating canine behavior. However………

Slide6: 

It is difficult to study the behavior of pet dogs Temperament tests are time consuming, and their reliability and validity are uncertain. Behavioral observations in the home environment are logistically difficult, and some behavioral responses are rare events. Clinical samples, and samples derived from shelter intake statistics, do not provide an accurate reflection of the frequency and distribution of behavioral traits in the general dog population (biased toward the most antisocial/ inconvenient traits). So we decided to opt for…………..

Slide7: 

An Indirect Method of Recording Behavior Using a Questionnaire Nobody knows more about a dog’s typical behavior than its owner/keeper. It is possible, by asking the right questions, to access that knowledge in a form that is accurate, quantitative and reliable. Makes Two Assumptions

Slide8: 

Advantages Makes data gathering relatively easy and inexpensive. Has a long and established track record in studies of ‘temperament’ in young children. Disadvantages Reliability and validity of data derived from questionnaire assessments of dogs is unknown. Viewed with skepticism by some (many?) in the field of canine behavior.

Slide9: 

Research goals: To design and develop a survey instrument (questionnaire)1 to access quantitative behavioral information on companion (pet) dogs. To determine the reliability and validity of this instrument. 1 The University of Pennsylvania Behavioral Assessment & Research Questionnaire (PennBARQ).

Slide10: 

PennBARQ Survey 152 questions:5-point visual analogue (semantic differential) rating scales (20-25 minutes to complete) Sociability (8 items) Training and Obedience (13) Aggression (44) Fear and Anxiety (22) Separation-related problems (15) Excitability (12) Attachment and Attention-seeking (9) Miscellaneous (29)

Slide11: 

PennBARQ Survey cont. Respondents: 960 regular dog owners (former vet hospital clients), 1000 breed club members; one dog/household 200 owners of ‘problem’ dogs Dogs Age:  1 year old (mean ±SE = 69±1 months) Breed: no restriction (n>100: Basset Hound, Dachshund, Golden Retriever, Rottweiler, Shetland Sheepdog, Siberian Husky, Yorkshire Terrier, West Highland White Terrier) Sex: no restriction (male:female = 1:1) Methods Factor analysis, reliability estimation & validation.

Slide12: 

Factor Analysis Q11 Q12 Condenses information by grouping items into related factors or subscales.

Slide13: 

Subscale composition Owner-directed aggression = Verbally corrected or punished. When toys, bones or other objects taken away by a member of the household. When bathed or groomed by a MoH. When approached directly by a MoH while eating. When his/her food is taken away by a MoH. When stared at directly by a MoH. When stepped over by a MoH When MoH retrieves food or objects stolen by dog Aggression (growls, barks, bares teeth, lunges, snaps, bites or attempts to bite) when:

Slide14: 

Items not included in the subscale (owner-directed aggression) When moved from, or disturbed at, a favorite resting or sleeping place. When physically corrected or punished by a MoH. When played with by a MoH. When having leash put on by a MoH. When pushed down/kept in a down position by a MoH. Suddenly or unpredictably toward a MoH for no apparent reason.

Slide15: 

The best surveys are: Reliable: measures are repeatable and consistent; free from random errors. (coefficient alpha  ≥ 0.7) Valid: they measure what they are designed to measure. Low High Low High

Slide16: 

Subscale Reliability Estimates (PennBARQ Survey: 11 factors, 69 variables, 57%, N=657)

Slide17: 

PennBARQ data from a sample of ‘problem’ dogs compared with independently-derived clinical diagnostic information on the same animals. The former obtained from a subset of owners who have sought treatment for their animals’ behavior problems. The latter provided by a cooperating group of clinical behavior specialists. Validation Procedure Methods: Owners are sent the PennBARQ to complete when they phone for their appointment with the behavior specialist whose diagnosis is then formulated subsequently and sent to us independently without reference to the completed questionnaire.

Slide18: 

Frequency of major behavioral problems (Source: Behavior Clinics, N=200)

Slide19: 

Comparison of PennBARQ scores for ‘problem’ dogs (upper edge of grey area) versus ‘normal’ dogs (lower edge). (N= 200, means + 95% CIs)

Slide20: 

Comparison of PennBARQ scores between breed club members, vet clients, and ‘problem’ dogs: (error bars = 95% confidence intervals)

Slide21: 

Conclusions: Except for “body sensitivity”, all subscales of the PennBARQ instrument have good reliability characteristics. Construct validity of 7 subscales were confirmed (little or no overlap between ‘problem’ and ‘normal’ dogs). Validity of 4 subscales - chasing, trainability, excitability, body sensitivity - still to be determined. With an appropriate survey instrument, reliable and valid measures of a dog’s behavior and temperament can be obtained from its owner/keeper/guardian.

Slide22: 

Can we detect consistent and meaningful breed differences in behavior using the PennBARQ instrument?

Slide23: 

Breed Comparisons (all data from breed club members)

Slide24: 

Breed Comparisons (all data from breed club members)

Slide25: 

Westies vs. seven other breeds

Slide26: 

Westies vs. seven other breeds

Slide27: 

Westies vs. seven other breeds

Slide28: 

Although still undergoing development, the PennBARQ instrument shows promise as: A valid and reliable framework for describing and evaluating behavior and behavior problems in dogs. A basis for selecting dogs for working and/or companionship roles. A method of screening dogs for behavior problems. A technique for assessing the behavioral effects of remedial ‘treatments’. A research tool for exploring the epidemiology, etiology, and ontogenesis of canine behavior and behavior problems.

Slide29: 

Acknowledgments The Arell Foundation The Kenneth A. Scott Charitable Trust The Pet Care Trust The University of Pennsylvania Research Foundation Sharon Crowell-Davis Benjamin Hart Suzanne Hetts Katherine Houpt Anne McBride Pamela Reid Various Breed clubs All participants