Does cooperating in a Black virtual body reduce implicit bias?
White participants who embodied a Black avatar and built a puzzle together with a Black virtual partner showed a larger drop in implicit racial bias than those who built it alone while the partner watched.
Source
Exploring the Effect of Cooperation in Reducing Implicit Racial Bias and Its Relationship With Dispositional Empathy and Political Attitudes
Study at a glance
- Design
- Human experiment — Randomised two-group VR experiment (cooperative vs neutral presence of a Black avatar partner) with pre/post IAT and closeness measures, plus baseline correlations with empathy and politics
- N
- N=40 · 40 White adults randomly split between the Cooperative and Neutral groups; the empathy and politics analyses pool all 40
- Population
- White adult volunteers in France (13 men, 27 women) embodied in a same-sex Black virtual body
- Outcome
- Change in racial IAT score from before to after VR; change in Inclusion of the Other in the Self (IOS) closeness; baseline IAT predicted by IRI empathy subscales and political orientation
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
White adults wore a VR headset and motion-capture suit so that they saw and moved a Black virtual body from a first-person view, including in a virtual mirror. Opposite them sat a Black avatar controlled by an experimenter. In the Cooperative group the two had to pass bricks to each other to complete a 3D Colosseum puzzle; in the Neutral group participants finished it alone while the avatar watched. Implicit racial bias (IAT) and felt closeness to the partner (IOS circles) were measured before and after about 9 minutes in VR, and empathy, social dominance and political orientation were measured by questionnaire beforehand.
What they found
Both groups felt ownership of the Black body and rated the session as similarly pleasant, but only the Cooperative group rated it as cooperative. After controlling for change in closeness, IAT scores fell in the Cooperative group (mean change about -0.11) but not in the Neutral group (about +0.02). In the Neutral group, people who came to feel closer to the partner showed larger bias reductions, whereas in the Cooperative group the drop did not depend on closeness. At baseline, higher empathic concern and perspective-taking went with lower implicit bias, higher personal distress with higher bias, and more left-wing participants scored higher on empathic concern.
The limits
What it doesn't show
With only 40 participants (about 20 per group) the effects are fragile, and the authors themselves frame the study as hypothesis-generating. There was no White-avatar or non-VR cooperation condition, so the study cannot say whether being in a Black body adds anything beyond cooperating in general. Participants knew the partner was really a White experimenter, and the IAT was re-taken minutes later, so nothing is known about lasting change or real-world behaviour. The empathy and politics findings are cross-sectional correlations in a small sample, with politics measured by a single self-rating item.
Key terms
- Implicit Association Test (IAT)
- A reaction-time task that infers automatic associations from how much faster people sort pairings such as White+good/Black+bad than the reverse pairings.
- Virtual body ownership
- The illusion that an avatar's body is your own, produced by seeing it from a first-person view while it moves in sync with your real movements.
- Inclusion of the Other in the Self (IOS)
- A single-item scale of overlapping circles on which people pick how much they feel another person overlaps with themselves, used as a measure of psychological closeness.
- Interpersonal Reactivity Index (IRI)
- A questionnaire of dispositional empathy with four subscales: perspective-taking, fantasy, empathic concern and personal distress.
- Contact hypothesis
- The idea that prejudice between groups is reduced by positive, cooperative contact under the right conditions.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What task did participants perform in virtual reality?
Common questions
Why compare cooperation to a 'neutral' partner rather than to no partner at all?
Both groups saw the same Black avatar and did the same puzzle, so the only difference was whether they had to work together; this isolates cooperation as the active ingredient rather than simply seeing another Black person.
Does this show that VR permanently reduces racism?
No. Bias was re-measured only minutes after a short session, on an implicit reaction-time measure, in 40 people; durability and effects on real behaviour were not tested.
Why include change in closeness in the analysis?
Earlier work suggested that liking the virtual partner drove bias reduction, so the authors controlled for it; cooperation still lowered bias, and closeness only mattered in the neutral group.
More on Prejudice & bias