YouTube videos help researchers study dog bites

Preventing dog bites is an increasingly important public health and political issue with implications for both human and animal health and welfare. However, it remains difficult for researchers to understand the circumstances leading up to dog bites, with most studies relying on evidence collected after bites happen, such as hospital records and victim interviews.

In a new study published in Scientific Reports researchers have, for the first time, used YouTube videos to directly observe and analyse dog bites in situ.

Lead author Sara Owczarczak-Garstecka said: "Online videos present us with an unexplored opportunity to observe dog bites first-hand, something which is just not possible using other methods. Making more use of this type of shared content for research could help us better understand how and why bites occur and contribute to the development of bite prevention strategies."

Using search terms such as "dog bite" and "dog attack" the researchers sampled 143 videos that were uploaded to YouTube between January 2016 and March 2017. For each video the context of bites, bite severity, victim and dog characteristics were recorded. For 56 of these videos they were also able to analyse the details of human and dog behaviour leading up to the bite.

The researchers acknowledge that YouTube videos of dog bites are likely subject to some bias, with, for example, bites by small dogs perhaps perceived as "comical" and therefore more likely to be uploaded online.

The findings reveal that despite this potential bias, the demographic characteristics of the victims and dogs seen in YouTube bite videos, such as breed type and victims' sex and age, are consistent with those found in previous studies. Common dog breeds observed included Chihuahuas, German Shepherds, Pit bulls and Labrador Retrievers. Around 7 in 10 of the bite victims in the videos were male, while more than half of bites observed were to children and infants.

Although this small study did not allow an exploration of the causal relationship between human behaviour and dog bites, some behaviours that have been previously observed within the context of dog bites were observed here to precede a bite. For example, the researchers observed that tactile contact with a dog increased approximately 20 seconds before a bite, as did standing or leaning over a dog.

Sara Owczarczak-Garstecka added: "These findings could offer some valuable new insight for the development of bite prevention strategies. Prevention messages could emphasise the risk of leaning over a dog and simply advise avoiding contact with a dog when possible or in doubt."

Future research plans to better understand people's behaviour around dogs and their perceptions of dog bites include a series of interviews with dog owners, people who work around dogs and bite recipients.