What caregivers don’t tell you ... A comparison between survey responses and home videos of cat-cat interactions

Domestic cats are increasingly popular as companion animals, but behavioural problems are often reported, especially in multicat households. Social tension is a common stressor, so understanding intercat interactions and their dynamics is crucial. Nevertheless, direct research in the home setting is rare. As caregivers witness their cats’ behaviour on a daily basis, they are a potentially important source of information, but might be unreliable and subject to bias. This study examined the reliability of caregiver reporting by comparing survey answers with behaviours observed in home videos collected after the survey was completed. The occurrence of five cat-cat interactions (head rubbing, allogrooming, sleeping in physical contact with each other, tail up greetings and social play) was examined in 42 two-cat households using 210 survey answers and 775 videos. The percentage of false negative survey responses for behaviours observed in the videos was conservatively estimated (cFN) at 8%, with 22.9% of the negative answers being falsely negative (FOR – False Omission Rate) and 77.1% truly negative (NPV – Negative Predictive Value). Broad false negatives (bFN), which included uncertain responses as negative reports, were 9.5% of the survey responses with a FOR of 75% and NPV of 25% in this context. Highest values were obtained for head rubbing (cFN: 10.5%, bFN: 14.3%) and allogrooming (cFN: 9.8%; bFN: 11.9%). When focusing on individual cat caregivers, 14 out of 42 caregivers (33.3%) failed to reliably report the occurrence of at least one of the surveyed cat-cat interactions. For interactions that were seen on camera, 23.8% of caregivers (10/42) responded that their cats did not show these interactions and 9.5% (4/42) reported uncertainty about whether it ever occurred. These results should be considered a lower estimate of the magnitude of errors (false negatives) in caregiver reports, and their implications need to be considered in both research that depends on caregiver report, and clinical assessments within behavioural medicine. Many cat-cat interactions, and in particular head rubbing and allogrooming, will be underreported when relying exclusively on caregiver reporting.