Assessment of the Agreement between Fosters and Trainers Evaluating Behavior with the C-BARQ and Behavior Checklist presented by Elizabeth Hare

Working dog programs have an interest in the early identification of puppies who are likely to succeed in training. The Penn Vet Working Dog Center begins training and fostering puppies at about eight weeks of age while collecting several types of behavioral data. The Canine Behavioral Assessment and Research Questionnaire (C-BARQ) contains 100 items grouped into 14 subscales and is usually completed by fosters. The Behavior Checklist (BCL) contains 44 items which partially overlap with C-BARQ items and subscales. After matching within dog and age group, there were 108 records on 83 dogs. Spearman correlations were calculated as a measure of agreement between the BCL item scores and C-BARQ subscales and individual items using the Holm correction for multiple testing. When testing C-BARQ subscales, BCL and C-BARQ stranger-directed aggression were moderately correlated (0.49, p < 0.001), as were C-BARQ dog-directed aggression and BCL stranger-directed aggression (0.42, p < 0.05). C-BARQ dog-directed aggression was correlated with BCL resource guarding around other dogs (0.46, p < 0.05). When testing BCL items with individual C-BARQ items, C-BARQ dog-directed aggression items were moderately but significantly correlated with seven BCL items including stranger-directed aggression, dog-directed aggression, and resource guarding around dogs. C-BARQ  stranger-directed aggression items were correlated with seven BCL items including fear and aggression toward strangers. The moderate correlations for fear and aggression and lack of correlation for other traits suggest that dogs’ behavior may be different in their foster homes than in training or that trainers evaluate dogs differently than fosters.