In a VR café, social anxiety - not attentional focus - predicted less stable speech in adults who stutter

Bauerly KR, Mefferd A · 2026 · Journal of Fluency Disorders · Experimental · n = 40 · Adults who stutter (n=17) vs adults who do not stutter (n=23), ages 18-45 · DOI
Evidence certainty: Low certainty
How this was rated

A single, small, single-site sample (17 adults who stutter, 23 controls) with a correlational/predictive design - so it shows associations, not cause. Several of the anxiety and attention questionnaires were newly created and, as the authors note, not yet reliability-tested. Balancing factors: the within-subject stress manipulation was physiologically validated (skin conductance rose significantly from the low- to high-stress scene), the analysis and assumptions are reported transparently, and it is NIH-funded, peer-reviewed work. Rated consistently with other single small studies in the Hub - directionally informative rather than definitive.

Ratings use a simplified four-tier scheme (High, Moderate, Low, Very Low) informed by the GRADE working group. Learn more about how studies are rated.

In a custom café built for the study with withVR software, higher social anxiety - but not where a person focused their attention - predicted more variable lip movements in adults who stutter speaking under social pressure. Seventeen adults who stutter and 23 who do not repeated sentences in a low- and a high-stress virtual café; the high-stress scene reliably raised everyone's physiological arousal, and in the stuttering group both self-reported anxiety and skin-conductance increases predicted greater articulatory variability, while self-reported attentional focus did not.

Clinical bottom line

A small, correlational study indicating that social anxiety - more than where speakers direct their attention - is what destabilizes speech movement in adults who stutter under social stress, and that a purpose-built VR social scene can reliably and controllably create that stress in the lab.

Key findings

  • Higher social anxiety - both self-reported and shown by a rise in skin conductance - predicted more variable lip movement in adults who stutter under social stress
  • Where speakers focused their attention (self vs. listener) did not significantly predict speech-movement stability in either group
  • Adults who stutter showed more variable articulation than controls; movement duration did not differ between groups
  • The custom VR café reliably raised physiological stress from the low- to the high-pressure scene in both groups, confirming it as a working social-stress paradigm
  • In non-stuttering adults, neither anxiety nor attention predicted articulatory variability

Background

Social anxiety is common in adults who stutter - far more common than in the general population - and it tends to spike in situations where a person feels evaluated. A separate line of research shows that heightened emotion can destabilize the fine motor control of speech. This study sat at the intersection of those two ideas, asking a specific question: when an adult who stutters speaks under social pressure, what actually drives their speech to become less stable - the anxiety itself, or a shift in where their attention is focused?

The distinction matters clinically. If anxiety is the driver, therapy should target the anxiety. If attentional focus is the driver, therapy might instead coach speakers on where to direct their attention while talking.

What the researchers did

Seventeen adults who stutter and 23 adults who do not (ages 18 to 45, spanning a wide range of social-anxiety levels) repeated two sentences inside a virtual-reality café, viewed through an HP Reverb G2 Omnicept headset. The café was a custom environment built for the study with withVR software; the same café is one of the situations in Therapy withVR.

Each person spoke in two conditions. In the low-stress scene they sat with a single friendly avatar. In the high-stress scene they faced a table of nine avatars - some smiling and nodding, some frowning and drumming their fingers, some neutral - in a noisy café. While they spoke, the researchers recorded skin conductance (a physiological marker of stress), lip and jaw movement (via 3D motion capture), and, afterward, self-reported anxiety and where the person had focused their attention. Crucially, and unlike the authors’ earlier work, participants were not told where to direct their attention - it was left to occur naturally.

The team then used regression to test which of these factors predicted two measures of speech-motor control: how long the movement lasted, and how variable the lip movement was from one repetition to the next.

What they found

The high-stress café did its job: skin conductance rose significantly from the low- to the high-stress scene in both groups, confirming the environment reliably produced social stress.

In the adults who stutter, higher social anxiety - captured both by self-report and by the rise in skin conductance - significantly predicted more variable lip movement. Where the speaker focused their attention did not. Movement duration was not predicted by any factor. Adults who stutter were more variable than the control group overall, but the two groups did not differ in how long their movements took. In the non-stuttering group, neither anxiety nor attention predicted articulatory variability.

In short: for adults who stutter speaking under social pressure, it was the anxiety - not the attentional focus - that tracked with less stable speech movement.

Why this matters

For clinicians, the practical takeaway is that lowering social anxiety may be a more direct route to steadier speech under pressure than coaching a speaker on where to place their attention. That lends support to therapy approaches - avoidance reduction, acceptance and commitment therapy, mindfulness - that work on the fear and avoidance surrounding speaking rather than on articulation alone.

For the wider field, the study is also a demonstration of VR as a research instrument. A purpose-built virtual café was able to produce graded, repeatable social stress - measurable in participants’ physiology - while high-precision motion capture recorded their speech. That combination is hard to achieve with real audiences, where the “dose” of social pressure cannot be standardized from one person to the next. It shows withVR software functioning as a controllable stimulus for tightly instrumented speech-motor experiments.

Limitations

This was a small, single-site study (17 adults who stutter), and its central finding is correlational - it links social anxiety to articulatory variability but cannot prove one causes the other. Several of the anxiety and attention questionnaires were created for the study and, as the authors acknowledge, have not yet undergone reliability testing; the attention measure may also have hit a ceiling. The speaking task - repeating fixed sentences - does not capture the demands of everyday conversation, and most participants stuttered in the mild range, so the sample may not represent the full spectrum. As with all lab studies of social stress, a virtual audience may not reach the intensity of a real one, though it trades some of that intensity for consistency and experimental control.

Implications for practice

The findings support prioritizing strategies that lower social anxiety - such as avoidance-reduction, acceptance, and mindfulness approaches - as a route to steadier speech under pressure, rather than cueing speakers to monitor their own articulation. Graded, controllable VR social situations, from a single friendly listener up to a mixed, reactive audience, offer a way to rehearse speaking under measured, repeatable social stress and to practice anxiety-management skills in context.

Implications for research

Demonstrates that a purpose-built VR social environment can generate graded, reproducible social-evaluative stress alongside concurrent kinematic (lip/jaw motion capture) and physiological (skin conductance) measurement - a practical paradigm for speech-motor and social-anxiety research. Replication with larger samples, everyday (non-repetition) speech, and reliability-tested attention measures is the clear next step.

Editorial notes from withVR

Where this connects to Therapy withVR

The study above is independent research and does not endorse any product. The notes below are commentary from withVR on how the themes in this research relate to features of Therapy withVR. The research findings are not claims about Therapy withVR.

Graded social situations (café and audience)

The study scaled social pressure from one friendly listener to a nine-person reactive audience in a noisy café. Therapy withVR is built around the same idea - starting a client in a low-pressure social situation and gradually increasing the demand as their confidence grows.

Avatars with varied reactions

The high-stress scene used avatars showing positive, negative, and neutral responses to create social-evaluative pressure - a way to expose clients to realistic, sometimes uncomfortable listener reactions in a safe, repeatable setting.

Cite this study

If you reference this study in your work, the canonical citation formats are:

APA 7th
Bauerly, K. R., & Mefferd, A. (2026). The role of social anxiety and attentional focus in a virtual reality environment on articulatory control in adults who stutter. Journal of Fluency Disorders. https://doi.org/10.1016/j.jfludis.2026.106236.
AMA 11th
Bauerly KR, Mefferd A. The role of social anxiety and attentional focus in a virtual reality environment on articulatory control in adults who stutter. Journal of Fluency Disorders. 2026. doi:10.1016/j.jfludis.2026.106236.
BibTeX
@article{bauerly2026,
  author = {Bauerly, K. R. and Mefferd, A.},
  title = {The role of social anxiety and attentional focus in a virtual reality environment on articulatory control in adults who stutter},
  journal = {Journal of Fluency Disorders},
  year = {2026},
  doi = {10.1016/j.jfludis.2026.106236},
  url = {https://withvr.app/evidence/studies/bauerly-2026}
}
RIS
TY  - JOUR
AU  - Bauerly, K. R.
AU  - Mefferd, A.
TI  - The role of social anxiety and attentional focus in a virtual reality environment on articulatory control in adults who stutter
JO  - Journal of Fluency Disorders
PY  - 2026
DO  - 10.1016/j.jfludis.2026.106236
UR  - https://withvr.app/evidence/studies/bauerly-2026
ER  - 

Know of research that should be in this hub? If a relevant peer-reviewed study is not listed here, send the reference to hello@withvr.app. The hub is kept up to date as the literature grows.

Funding & independence

This study used a custom virtual-reality environment built for the researchers with withVR software. The research itself was designed, conducted, analyzed, and authored independently by Bauerly and Mefferd and was supported by a US National Institutes of Health exploratory grant (R21CC020013); withVR did not fund, author, or influence the study's design, analysis, or findings. The authors declare no conflict of interest - see the publication for their own disclosures.

Last reviewed: 2026-07-03 Next review due: 2027-07-03 Reviewed by: Gareth Walkom