The experiment aimed to investigate the owner-reported anecdote that dogs sometimes greet owners displaying guilty behavior. Owners claim to be unaware of a dog's misdeed and assert it is the guilty behavior that informs them of the dog's infraction. We studied whether dogs that were disobedient in owners鈥?absences showed associated behaviors of guilt (ABs) upon owners鈥?return to a room. We also assessed whether owners could determine their dog's disobedience by relying solely on the dog's greeting behavior.
Behavioral analysis revealed no significant difference between obedient and disobedient dogs in their display of ABs after having the opportunity to break a rule in owners鈥?absences. Analyses at the individual level, however, revealed a significant increase in cross situational presentation of ABs only by dogs that transgressed in owners鈥?absences. While owners appeared able to determine whether or not their dogs ate in their absences, a subset of owners鈥攖hose whose decisions were most likely based solely on dog greeting behavior and not earlier experiment-generated cues鈥攚ere not better than chance in their determinations. Taken together, our findings suggest that dog presentation of ABs during greetings is not necessarily a reliable indicator whether or not a dog engaged in a misdeed. The investigated phenomenon appears to be very sensitive to the social condition, which includes owner prior experience with their dog in specific contexts.