Not By Simulation Alone (Or: Mo’ VR, Mo’ Problems…Unless You Do It Right)

We’ve often seen that old familiar question: “Do media technologies make it better or worse?” It’s been examined by many, particularly in general terms such as at the societal level (such as Neil Postman’s famous lack of love for television and electronic entertainment) and at the individual level over a broad range of time (such as the classic Carnegie Mellon HomeNet study on the effects of Internet use).

The answer to that “better or worse” question seems to often be: “Well, that depends.” I want to share a recently-published study that I believe is useful because it provides a very specific and important example of how precarious the tightrope can be between a technology helping and harming depending on how the technology is employed.

I had the opportunity to be involved in the study, a full report of which is in a recent issue of the Journal of Nervous and Mental Disease, but only in a limited role. Its leaders were Sriram Kalyanaraman and David L. Penn at the University of North Carolina, and Abigail Judge of UNC was also a co-author.

An image from the simulator.

The study examined the effectiveness of a virtual reality simulator provided by Janssen Pharmaceutica in inducing empathy toward persons with schizophrenia. Participants in the study were assigned to one of four conditions: 1) completing a written “empathy” task where they imagined what it would be like to suffer from schizophrenia and wrote about it, 2) using the virtual reality simulator to “experience” the viewpoint of a person with schizophrenia at a supermarket, 3) completing both the written task and the VR simulator, or 4) doing neither the written task nor the simulator. Participants in all of these conditions then answered questions assessing their perceptions toward people with schizophrenia, including feelings of empathy (such as how ““Disgusted,” “Scared,” and “Compassionate” they felt toward people with schizophrenia), feelings of social distance (such as “How would you feel having someone with schizophrenia as a neighbor?”), and attitudes toward persons with schizophrenia, among other measures.

Results for empathy and attitudes were simple: Scores were lowest in the control condition, slightly higher in the written empathy condition, still higher in the simulator condition, and clearly highest in the condition where participants completed both the written empathy and simulator tasks. Hooray for technology, right? Things got more complex with the data for social distance, which produced similar scores in all conditions except for the simulator-only condition, which produced significantly greater feelings of social distance (which isn’t good).

In other words, the VR simulator induced more empathy and better attitudes toward persons with schizophrenia, especially when accompanied by the written task … but the VR simulation without the written task made people more uncomfortable about inteacting closely with people who suffer from schizophrenia. The simulator was not only more effective, then, in reducing stigma when accompanied by the empathy task; the simulator was actually counterproductive in some ways without the empathy task.

Although we can’t say whether a similar pattern would be observed in other contexts, it may well provide a useful caution for anyone interested in the potential of high-tech simulations: Technologies like VR simulations can be powerful tools in the right context, but by themselves may do as much harm as good (or more).