Three validation studies (n=40) - stuttering behavior, anxiety, and cortisol in VR mirrored real speaking situations

Background

A persistent gap exists between what happens in the speech clinic and what happens in everyday life. People who stutter often communicate fluently with their clinician but face significant challenges the moment they step into a meeting, a classroom, or a social gathering. Brundage argued that VR could serve as a middle ground - an environment realistic enough to provoke genuine communicative responses, yet safe and controllable enough to allow systematic practice.

What the researcher did

This tutorial paper brought together data from three separate validation studies conducted at the Stuttering Research Laboratory at George Washington University, in collaboration with Virtually Better, Inc. (the commercial VR-software vendor):

• Study 1 (Brundage et al., 2006) - twenty adults who stutter completed both a supportive and a challenging virtual job interview, with interview order and interviewer gender counterbalanced. Stuttering frequency was the primary outcome.
• Study 2 (Brundage et al., 2007) - ten adults who stutter and ten adults who do not stutter each gave three five-minute extemporaneous speeches: one to a live audience, one to a challenging virtual audience (~20 inattentive listeners), and one to a neutral virtual audience. Speakers completed self-report measures of communication apprehension (PRCA), confidence (PRCS), and self-perceived communication competence (SPCC) before and after each speech.
• Study 3 (Duncko, Brundage, Graap, Kling, & Gold, 2006) - salivary cortisol was measured in adults who stutter during virtual job interviews and correlated with stuttering severity and self-rated communication apprehension.

Study 2 was the only one of the three to include both PWS and non-stuttering participants; Studies 1 and 3 were conducted with adults who stutter only.

What they found

The combined evidence was consistent across behavioral, psychological, and physiological levels. In Study 1, stuttering frequency was significantly higher during the challenging virtual job interview than the supportive one (one-tailed t = 2.14, p = .02), demonstrating that virtual social challenge could elicit increases in disfluency. In Study 2, comparisons between virtual and live audience conditions revealed no significant differences in self-reported communication apprehension, confidence, or self-perceived competence; percent syllables stuttered did not differ significantly between the virtual and live audience conditions either. In Study 3, salivary cortisol correlated positively with stuttering severity and with self-rated communication apprehension during the virtual job interviews, indicating that the body’s stress response was genuinely engaged. Participants in the debriefs commonly described realistic urges to avoid words, tension in their speech muscles, and fear of negative evaluation - several reached out to shake the virtual interviewer’s hand.

Why this matters

By demonstrating convergence across multiple measurement channels, this paper established a strong foundation for VR as a legitimate tool in stuttering services. The concept of VR as a bridge - more demanding than the therapy room but safer than the real world - provides a practical framework for clinicians designing graded speaking hierarchies.

Limitations

The three studies were conducted with relatively small samples and were combined post hoc rather than designed as a single investigation. The populations were predominantly adults, leaving questions about whether similar patterns would emerge with younger people who stutter. The paper presented a theoretical framework but did not include longitudinal data on whether VR-based practice led to lasting communication gains.