Cat Play or Cat Fray? Investigating the Accuracy of Veterinary Student Classifications of Human-Cat Interactions

Abstract Accurate interpretation of domestic cat behavioural cues is critical for ensuring animal welfare and veterinary staff safety in clinical settings; however, the ability of veterinary students to reliably distinguish between affiliative and agonistic human-cat interactions remains underexplored. This study evaluated the ability of Australian veterinary students to accurately determine feline behavioural cues during human-cat interactions, utilising diagnostic sensitivity and specificity as measures of behavioural interpretation accuracy. A cross-sectional survey was distributed online to Australian veterinary students enrolled in an accredited veterinary program. Participants viewed 10 of 16 randomly selected short videos of human-cat interactions. The videos utilised were previously classified by an expert panel (of five) as affiliative or agonistic interactions, based on ≥80% consensus between the experts. Students classified each interaction as affiliative or agonistic , and diagnostic sensitivity (DSe) and diagnostic specificity (DSp) were calculated to determine student accuracy. Univariate metaregression was used to assess possible associations between accuracy measures and cat vocational experience, confidence in rating, cat ownership, and university year level. Students demonstrated higher sensitivity (DSe; 86.9%; 95% CI 84.3%–89.2%) than specificity (DSp; 65.4%; 95% CI 61.8%–68.8%). Student accuracy decreased when assessing subtle videos (DSe 78.5%; 95% CI 74.0%–82.4% and DSp 41.3%; 95% CI 35.9%–46.9%), suggesting that subtle agonistic cues were frequently misclassified. There were no significant associations identified between diagnostic accuracy and any predictor (all p >0.05). Australian veterinary students may require additional education on cat behaviour cues, particularly subtle agonistic behaviours.