VR@Manchester

VR@Manchester Innovation Club - February 2021

<p>Human Interpretation and Behaviour in the Presence of a Virtual Reality Dog Model</p><p>James Oxley, University of Liverpool</p>

26th February 2021 12:00 — 13:00

Online - Zoom


Watch the edited Presentation Here.

Abstract: A pilot study investigating human interpretation and behaviour in the presence of a virtual reality dog model (Dog Assisted Virtual Environment) displaying aggressive and unresponsive behaviours.

Introduction: Dog bites are a serious public health matter that negatively impacts the health and welfare of dogs and humans. However, limited research has been conducted on the assessment of human behaviour with a real aggressive dog due to the risk of physical and/or psychological damage. This study aims to assess and interpret human behaviour towards an interactive virtual dog displaying a range of aggressive behaviour.

Methodology: An interactive virtual reality (VR) dog model was developed for the HTC Vive Pro with the assistance of behavioural experts. Two interactive modes were implemented i) Aggression: behaviours based on the canine ladder of aggression and dictated by user movement/location; ii) Unresponsive: passive behaviours (stand, sit, sniffs the ground, lick lips, head moves) not dictated by participants movement/location. Sixteen university students took part in two five-minute tasks and were randomly assigned to two groups starting with either the unresponsive or aggressive dog first. Participants were instructed to explore an area (6x2m) resembling a living room. Participants starting position was 4.6m away from the closest point to the dog. Coordinates of the dog and participant were recorded (5hz). Comparisons between the closest distance a participant got to the dog between scenarios was analysed per group using a Wilcoxon signed-rank test. A presence questionnaire recorded participants feeling of ‘being there’.

Main results: Participants head and hands moved significantly closer to the unresponsive dog compared to the aggressive model for both groups (p<0.05). Three participants were ’bitten’ by the dog. Participant feedback and mean total presence scores indicated the experience was realistic (aggressive (159/203) unresponsive (149/203)). Body movements and perceived emotional/motivational aspects were most often reported when describing behaviours.

Conclusion: This unique VR dog model displays aggressive behaviours which can be used to understand human behaviour around dogs. Participants stayed further away from an aggressive dog in VR.