Stuttering is a complex, multidimensional disorder involving affective, behavioral and cognitive components that influence a person's quality of life (Guitar, 2006; Smith & Kelly, 1997; Yaruss & Quesal, 2004; Yaruss, Quesal & Molt, 2004). Virtual Reality (VR) is a tool that has improved the treatment of simpler disorders that involve one or more affective, behavioral, or cognitive components. My colleagues and I are exploring how VR can be applied to stuttering. If it can be shown that VR elicits responses similar to those in the real world, then the next step will be determining how to integrate VR into stuttering assessment and treatment. The purposes of this paper are to: 1) define virtual reality, 2) describe how VR has been used to successfully augment the treatment of disorders other than stuttering, 3) summarize the findings from my research on applications of VR to stuttering, and 4) discuss some potential challenges in research on stuttering.

What is Virtual Reality?

Virtual reality (VR) is a human computer interaction in which users are active participants in a computer generated three dimensional world (Schultheis, Himelstein, & Rizzo, 2002). Many people have interacted in virtual worlds when playing video games. When VR is used in rehabilitation, clients interact with virtual people that the treating clinician can control to varying degrees. In order to feel immersed in the virtual world, clients may wear head-mounted displays (specialized optics and trackers that allow them to see different parts of the virtual world when they move their heads), haptic gloves (gloves that allow for the perception of force against objects in the virtual world), and noise-canceling headphones (to decrease auditory distractions). Various scents (coffee in the coffee maker, the secretary's perfume, the paint in the custodial closet) added to the virtual world increase the sense of presence in the virtual environment.

VR offers advantages in the treatment of various phobias and fears. First, practice in VR is safe. For example, it is much safer to desensitize oneself to a fear of heights on a virtual bridge rather than a real one. Second, virtual people react in predictable ways. Clinicians control the extent to which client's are asked to confront their fears and/or practice treatment strategies. In addition, these interactions are repeatable and can be arranged hierarchically to suit individual needs, allowing for systematic practice and feedback from the treating professional. Third, clients do not leave the treatment room and do not interact with real people, thereby assuring confidentiality. Finally, VR may inspire clients to higher levels of motivation and participation when practicing treatment techniques compared to role-playing with clinicians (Graybiel, 2004).

VR and the treatment of anxiety, fear, motor learning and body image

Stuttering involves affective (e.g., fear of public speaking, speech-related anxiety), behavioral (e.g., sound/syllable repetitions, avoidance behaviors) and cognitive processes (e.g., awareness of one's stuttering, knowledge of stuttering). Affective, behavioral, and cognitive aspects of various disorders have been successfully treated using VR-augmented treatment.

The earliest VR-exposure studies used virtual elevators, bridges, and balconies to assist in the treatment of fear of heights (Rothbaum, Hodges, Kooper, Opdyke, Williford, & North, 1995a, 1995b). Persons participating in these studies were able to reduce their fear of heights while in VR and were able to extend their reduced fear response to real world settings (Rothbaum, Hodges, Smith, Lee, & Price, 2000; for comprehensive reviews see Glantz, Rizzo, & Graap, 2003; Krijn, Emmelkamp, Olafsson & Biemond, 2004; Riva, 2005). Research on the fear of inanimate objects or situations, such as spiders or open areas has shown that VR based exposure treatments work as well as real world exposure treatments in reducing fear (Rothbaum, et al., 2000; Rothbaum, Anderson, Zimand, Hodges, Lang, & Wilson, 2006).

VR has been used to improve motor behaviors, such as increasing range of motion in the arms and hands in stroke survivors (Merians, Jack, Bonian, Tremaine, Burdea, Adamovich, Recce, & Poizner, 2002; Holden, 2005; Sisto, Forrest, & Glendinning, 2002). VR has also been used to retrain behaviors associated with activities of daily living post brain injury (Christiansen, Abreu, Ottenbacher, Huffman, Masel, & Culpepper, 1998; Albani, Pignatti, Bertella, Priano, Semenza, Molinari, Riva, & Mauro, 2002; Zhang, Abreu, Masel, Scheibel, Christiansen, Huddleston, & Ottenbacher, 2001). Practicing motor skills in environments that have meaning for learners improves motor learning, motivation and generalization of new skills to novel environments (Graybiel, 2004; Duffy, 1995; Robin, Toussaint, Blandin & Proteau, 2005; Stokes & Baer, 1977; Ingham, 1984). Unfortunately, it is often not efficient or feasible for clinicians to escort clients to meaningful venues outside the clinic, and it is difficult to create situations that routinely induce fear, anxiety and stuttering within the clinic. For example, most persons who stutter fear speaking in front of audiences but it is impractical to assemble audiences for each client. Similarly, it may be awkward to have one's therapist along for a job interview.

VR has also been used with anorexic persons to assist them in changing their internal images of their bodies (Riva et al., 1998). Accurate self-evaluation of body image is essentially a cognitive process.

In terms of activities that involve communication interactions, VR has been used successfully in studies involving fear related to interpersonal interactions such as social phobia and the fear of public speaking (Anderson, Zimand, Hodges, & Rothbaum, 2005; Harris, Kemmerling, & North, 2002; Pertraub, Slater, & Barker, 2001). VR has not been used to assess or treat speech motor skills such as techniques that one might reasonably learn in during treatment for stuttering (e.g., pull-outs, cancellations, easy onsets, etc), or to address speaking fears associated with stuttering. The necessary first step before VR is used in treatment is to evaluate the feasibility and validity of VR environments for potential use in stuttering treatment. This is the work that is currently on-going in my research lab.

VR and stuttering

My research has focused on the feasibility of using VR with persons who stutter (PWS), and on validating VR's use with PWS. These studies are precursors to larger treatment studies using VR to augment treatment in stuttering.

In terms of feasibility, we needed to know if stuttering could be manipulated in VR; if it could not, then the treatment applications were a bit less promising. We knew from the research literature and from PWS self-reports that the following types of interactions tended to increase stuttering frequency during real world interactions: a) talking to persons in authority, b) talking under time pressure, and c) talking on the phone. Thus, to test the feasibility of VR for use in stuttering treatment, we set out to see if we could get similar findings in virtual environments. In conjunction with Virtually Better, Inc we developed four virtual job interview environments that systematically manipulated the level of authority and amount of time pressure in conversations with and without a telephone (Brundage & Graap, 2005). We found that PWSs stuttered more in virtual environments that involved speaking to persons in authority, under time pressure and on the phone; these findings suggest that it is feasible to manipulate stuttering in VR and that there is potential for use as an augmentation to treatment interactions (Brundage, Gibbons, Graap, Ferrer, & Brooks, in press).

Our second question had to do with the validity of VR for use with PWS. To answer this question we carried out a direct comparison of stuttering and related behaviors in real and virtual environments (Brundage, Hancock, Kiselewich, & Stallings, 2006). Ten PWS and 10 non-PWS gave 5-minute speeches to virtual and live audiences. Immediately prior to giving each speech, participants randomly selected a topic from a set provided by the investigator. One speech was delivered to a challenging virtual audience (Virtual Audience software, Virtually Better, Inc.) which consisted of approximately 20 listeners who did not pay attention, did not make eye contact, and made other non-verbal distracting behaviors (e.g., falling asleep). A second speech was given to a Live audience consisted of approximately 10 people; these people were not instructed on how to respond to speeches. Self-reports of speaking apprehension and confidence, as well as behavioral measures of stuttering frequency, were obtained after each speech.

Measures of speaking apprehension and confidence. There were no significant group differences in self-ratings of speaking confidence. The two groups did differ on measures of speaking apprehension, with PWS reporting higher apprehension than non-PWS.

When comparing VR versus Live Audiences, measures of speaking apprehension and confidence were similar for both types of audiences.

Measures of stuttering frequency. We calculated the percentage of stuttered syllables in each speech for the PWS group only. There were no significant differences in the %SS across audience conditions, suggesting that the frequency of stuttering is similar in virtual and real world conditions. These findings suggest that similar responses occur after speeches to virtual and live audiences.

Challenges and opportunities in research

Stuttering is a complex, multidimensional disorder involving affective, behavioral and cognitive components (Guitar, 2006; Smith & Kelly, 1997; Yaruss, Quesal & Molt, 2004). Stuttering is also a relatively rare disorder. Further complicating the research endeavor, researchers often want to "control" for related factors (such as severity, gender, amount and type of previous treatment), thus making it difficult to find a large number of participants who meet all criteria for research studies. In order to answer important questions regarding the nature, assessment and treatment of stuttering, it will likely be necessary for researchers to develop a database of research data on stuttering to share with each other, in order to "pool" information to make meaningful conclusions. There is precedent for this type of database; the Child Language Data Exchange System (CHILDES; http://childes.psy.cmu.edu) was originally developed to analyze and compare the conversational language of children. The CHILDES system is available free of charge and has a set of coding conventions (CHAT) and a computer-assisted language analysis program (CLAN) associated with it. CHILDES has been adapted for use with adults, including adults who stutter (Bernstein Ratner, 2004), and a set of coding conventions for coding disfluencies has been developed (see MacWhinney, 2006). CHILDES includes data sets of coded transcripts from many research studies. After research participants provide consent their transcripts are put into the CHILDES databank for use by other researchers. In this way, data accumulates and can be used to answer many questions about stuttering, child language disorders, or other disorders. Nan Ratner and her colleagues have been lobbying for a stuttering databank in CHILDES for years now, and I'd like to suggest that there is no time like the present to start this conversation.

Pooling data will require some consistency in assessment protocols so that data can be compared across studies, an old problem that needs to be revisited (St. Louis, 2006), and perhaps virtual reality can assist us in this. VR has potential for use as a tool to standardize assessment of communication interactions. If researchers, PWS and other interested parties do not come up with an assessment protocol that suits us, you can bet that someone else will, and we might not like that one nearly as much as one that we develop together.

Selected References

Brundage, S. & Graap, (2005). The Virtual Job Interview.[software]. (Available from Virtually Better, Inc., 2450 Lawrenceville Hwy, Suite 101, Decatur, GA 30033).

Brundage, S., Graap, K., Gibbons, K., Ferrer, M., & Brooks, J. (in press). Frequency of stuttering during challenging and supportive virtual reality job interviews. Journal of Fluency Disorders.

Brundage, S., Hancock, A., Kiselewich, K., Graap, K., Brooks, J., & Ferrer, M. (2006). PWS and Non-PWS self-reports of communication apprehension and confidence when giving speeches to virtual and live audiences. Paper submitted to the Proceedings of the 5th World congress on Fluency Disorders.

Glantz, K., Rizzo, A., & Graap, K. (2003). Virtual reality for psychotherapy: Current reality and future possibilities. Psychotherapy, 40, 55-67.

Holden, M. (2005). Virtual environments for motor rehabilitation: A review. Cyberpsychology and Behavior, 8, 187-219.

Krijn, M., Emmelcamp, P., Olafsson, R., & Biemond, R. (2004). Virtual reality exposure therapy of anxiety disorders: A review. Clinical Psychology Review, 24, 259-281.

Riva, G. (2005). Virtual reality and psychotherapy: Review. CyberPsychology and Behavior, 8, 220-239.

Riva, G., Bacchetta, M., Baruffi, M., Rinaldi, S., & Molinari, E. (1998). Experiential cognitive therapy: a VR based approach for the assessment and treatment of eating disorders. Stud Health Technol Inform. 1998;58:120-35.

St. Louis, K. (2006). Measurement issues in fluency disorders. In N. Bernstein Ratner and J. Tetnowski (Eds.). Current issues in stuttering research and practice. Mahwah, NJ: LEA.

Yaruss, J. S., & Quesal, R. (2004). Stuttering and the International Classification of Functioning, Disability and Health (ICF): An update. Journal of Communication Disorders, 37, 35-52.

You can post Questions/comments about the above paper to the author before October 22, 2006.