Virtual reality won鈥檛 magically cure stigma 鈥� and sometimes it makes things worse
Primary page content
For years,virtual reality has been sold as an “empathy machine”. Put on a headset, step into someone else’s shoes, and come out a better person. Professor Sylvia Xueni Pan has spent much of her career testing that promise. Her latest work on schizophrenia suggests the hype badly needs qualifying.
The interior of a place of safety unit recreated using VR
Pan and her collaborators - Jiaqi Yin and Shao-Wen Lee - built a detailed virtual re-creation of a鈥�減lace of safety鈥濃�痷nit in a Midlands hospital, where police bring people in acute psychosis for assessment. Using motion captured performances by trained actors they staged a ten minute consultation between a psychiatrist and an adolescent patient with schizophrenia.
Volunteers, mostly psychology students, experienced this consultation in VR. Crucially, they were randomly assigned to one of three viewpoints:鈥痩ooking on as a third person observer,鈥痚mbodying the doctor, or鈥痚mbodying the patient. Before and after the session, they completed a battery of questionnaires measuring鈥痚mpathy,鈥痚xplicit stigma鈥痑nd鈥痓ehavioural intentions鈥痵uch as willingness to accept someone with schizophrenia as a neighbour or marriage partner.
The headline result seems, at first glance, reassuring. 鈥淲e did find that people reported鈥痬ore empathy鈥痑fterwards,鈥� Pan says 鈥� consistent with a large body of VR research. But looked at more closely and the picture darkens.
On several scales, including鈥痯erceived recovery鈥痑nd鈥痵ocial restrictions, stigma increased鈥痑fter the VR consultation. Participants were more inclined to see schizophrenia as hopeless and in need of tighter control, even as they claimed to understand patients better.
This, for Pan, confirms a longstanding unease about 鈥渆mpathy VR鈥�. Self reported empathy, she argues, is a鈥痵oft target: after any vivid experience, people will tell you they can imagine the other person鈥檚 perspective.
That doesn’t mean they are less afraid, or more willing to share their lives with them.
Sylvia Pan, Professor of Virtual Reality
The most striking finding came from the鈥痓ehaviour intention鈥痬easures 鈥� the questions that matter most to people鈥檚 everyday lives. Here, there was a strong鈥痠nteraction effect鈥痙epending on which body participants inhabited in VR.
Those who simply鈥痺atched from the sidelines鈥痗ame out with鈥痟igher stigma: greater social distance, more restrictive attitudes. Those who鈥痚mbodied the patient鈥痵howed little overall change. But those who鈥痓ecame the doctor showed a鈥痳eduction in stigma鈥痮n some key behavioural items.
鈥淚n a way, the first two results are about VR in general,鈥� Pan explains. 鈥淵ou get more empathy, but also more bias. The鈥痚mbodiment result鈥痠s the interesting one: putting people in the shoes of the effective professional in a successful consultation seems to nudge them towards less stigma.鈥�
An implicit attitude test (IAT) tailored to mental health failed to show a clear shift, but Pan is cautious about interpreting that null result. Unlike race, schizophrenia is hard to represent visually, and the task design may not yet be sensitive enough to detect subtle changes.
What her study does show, she insists, is that鈥疺R is a double-edged sword. If you faithfully portray a distressing, disorganised psychotic episode, 鈥測ou may increase people鈥檚 sense of how frightening and unpredictable it is,鈥� Pan observed, and thereby worsen stigma 鈥� particularly among groups, such as nurses, who already face the brunt of violent or chaotic behaviour on wards.
A related project with longtime collaborator鈥疨rofessor Mel Slater鈥痑t the University of Barcelona underlines the point. There, participants 鈥渢ime travelled鈥� through their own life in VR under climate change: as a child in a greener past, a young adult in an anxious present, a parent in a damaged near future, and an older adult in a brighter world achieved through collective action. Those who鈥痚mbodied themselves鈥痑t each stage showed a鈥痵hort term reduction in carbon footprint鈥痶wo weeks later; those who merely observed did not. By six weeks, however, the effect had faded.
Across both studies, a common lesson emerges:鈥疺R doesn鈥檛 automatically make people more virtuous. It鈥痑mplifies whatever script you build鈥�� the role you assign the user, the tone of the encounter, whether the story ends in paralysis or agency.
Pan鈥檚 message to policymakers and practitioners is not to abandon VR, but to鈥痷se it with care鈥痑nd in collaboration. 鈥淚t鈥檚 a powerful tool,鈥� she says. 鈥淏ut you can鈥檛 just throw VR at stigma or climate change and assume it will help. You have to design the experience around鈥痗ompetence, agency and realistic hope鈥�� and you have to keep testing that it鈥檚 doing what you think it鈥檚 doing.鈥�