Reliability of assessment of dogs' behavioural responses by staff working at a welfare charity in the UK
Forty members of staff at Dogs Trust, a UK dog welfare charity, were asked to assess 20 dogs based on the dogs’ behaviour recorded on video. They were then asked to repeat these assessments 2 months after they had completed the first assessments using the same video. The data were then analysed for inter-rater reproducibility and intra-rater repeatability using the weighted kappa statistic. It was found that there was a moderate level of agreement between staff in their assessment of the dogs’ behavioural responses to a person ‘approaching their kennel’ and a poor to moderate level of agreement for behavioural responses to ‘general handling and grooming’. Additionally, there was only a low-level of agreement for many of the responses to a dog ‘meeting another dog’ and this may have been due to the small number of dogs that were able to be assessed in this situation. The inter-rater reproducibility was greater when the analysis was restricted to those members of staff who had formal training or more than 8 years of experience. The intra-rater repeatability was generally moderate to high in all cases indicating a good consistency by members of staff over time.
The current study shows that there is an overall moderate level of agreement amongst staff but as they gain experience and training their level of agreement tends to increase. This study also highlighted the need for the standardisation of the assessments currently used. It is important that staff are able to consistently identify behavioural characteristics of dogs in order to provide the best chance of a successful adoption.
Title: Reliability of assessment of dogs’ behavioural responses by staff working at a welfare charity in the UK
Authors: Gillian Diesel, David Brodbelt and Dirk U. Pfeiffer