Assessment of Welfare of Shelter Cats and Dogs: Inter-Observer Reliability Using the Penn Vet Quality of Life Tool App
Abstract
To increase positive shelter animal welfare, we must find the causes of poor animal welfare on an individual level by using a reliable and valid assessment. In order to analyze inter-observer reliability, two raters of different backgrounds simultaneously assessed the welfare of 80 randomly selected animals (40 cats and 40 dogs) in a US shelter using the Penn Vet Quality of Life Tool App. Data collection occurred one to three times a week between the hours of 8 am and 2 pm from December 2021 to March 2022. Environmental variables showed good agreement, most physical health variables showed moderate agreement (e.g., fecal scoring and dog coat condition) and some behavioral variables (e.g., fear and aggression) frequently showed poor agreement. The poor level of reliability may be driven by factors such as the rater’s experience and understanding of animal behavior, their familiarity with the shelter animals, familiarity with the app, or differences in the rater’s gender. To address these challenges and better understand the factors that impact the reliability of animal welfare assessments, we recommend future studies: (1) train raters before data collection, (2) recruit more raters, and (3) record animals during live assessments for later review.