One of the most socially devastating outcomes during the development of those with Autism Spectrum Disorder (ASD) is not learning social skills that neurotypicals (NTs) experience through hidden curriculum, yet these skills are difficult to teach in a meaningful way since those with ASD tend to lack the ability to generalize lessons from traditional interventions and the fear of failing in the real world is prohibitive to teaching through doing. Recent research has demonstrated that Virtual Reality (VR) may be able to overcome these barriers by allowing those with ASD to learn and generalize material in a safe and novel way, yet much work needs to be done in determining the efficacy of VR at teaching social skills.
Thus in this present study we investigated the effectiveness of incorporating traditional social skill interventions with VR. We created an automated virtual environment (VE), which uses social narratives to teach hidden curriculum in a non-invasive and simple way. To evaluate the effectiveness of a social skill-teaching tool with a PVE on a standard desktop display, we needed to answer two questions:
1) Can students with ASD acquire social skills from a virtualized intervention?
2) How much more effective, if any, is a virtualized intervention compared to a traditional intervention?
To answer the first question, we conducted a within-group study for students using our intervention. The evaluation was done with a pretest and posttest that tested social skill ability. If results from the posttest showed a significant improvement over the pretest, the students must of accepted the VE and learned from the intervention. Further, we added questions on the posttest to determine the ability to generalize the specific social skills taught with the intervention.
To answer the second question, we conducted a between-group study comparing social skill learning between two different groups of students with ASD. The control group experienced a traditional intervention and the second group experienced the VR intervention. The social skill learning was measured using a written posttest.
The participants in our study demonstrated the learning of social skills by using a virtualized school environment, indicating that a virtualized intervention is effective at teaching social skills. Likewise, individual users reported a deep sense of presence within the environment, demonstrating the ability of VR to create a meaningful environment for future learning/application that may be used as a safe and controlled alternative to real-world learning. Further, the students generalized social skill development beyond the VR setting indicating the ability to transfer knowledge outside of the VR experience.
In conclusion this work has demonstrated, through a human-based empirical study, that students with ASD not only are able to learn hidden curriculum-based social skills through VR, but VR is more effective than traditional interventions at teaching these social skills. This work has immediate practical use, as we plan on releasing the VR intervention soon. Additionally, the approach we took on virtualizing social narratives can be applied to other interventions, which may chave broad implications on future assistive technology.