Glossary

Acoustic voice parameters

Measurable features of the voice signal, including fundamental frequency (f0), intensity, jitter, shimmer, and pitch variability. Used in voice research and in studies of speaking under pressure.

Anticipatory anxiety

Anxiety that arises before a speaking situation begins. Often measured separately from anxiety during or after speaking.

Aphasia

An acquired communication difference, most often following stroke, that can affect speaking, understanding, reading, and writing. A recurring topic in Evidence Hub studies on aphasia and cognitive-communication after brain injury.

Apraxia of speech

A motor speech difference in which the planning of speech movements is disrupted. Recurring topic in cognitive-communication after brain injury research.

Avatar / agent

A virtual character inside a VR environment. An avatar is human-controlled (representing the user); an agent is software-controlled (driven by an AI or scripted system). Both terms refer to the same kind of on-screen character but answer different questions about who is acting through it. In the research literature, avatar behavior, realism, and variability are often studied variables.

Behavioral Avoidance Test (BAT)

A behavioral measure in which the participant is asked to approach or engage with a feared situation. Distance approached, time spent, and self-reported distress are recorded. Used in exposure-therapy research as an outcome measure that complements self-report scales.

Between-subjects design

A study design in which different participants are assigned to different conditions. Contrasts with within-subjects, where every participant experiences every condition.

Blinding

Keeping participants, researchers, or assessors unaware of who received which condition, to reduce bias. Single-blind, double-blind, and triple-blind describe how many parties are blinded.

Communicative realism

The realism of a virtual character's communicative behavior - how the character moves, gestures, looks back, and responds in real time. In the research literature, communicative realism (movement and behavior) is generally found to matter more than photorealism (how the character looks) for producing realistic communicative responses in the user.

Condition (in research)

One of the experimental settings a participant is exposed to. A study comparing real audiences with virtual audiences has (at least) a real-audience condition and a virtual-audience condition.

Continuous Performance Task (CPT)

An attention-and-vigilance paradigm in which the participant responds to specified target stimuli within a longer stream. Used in research on attention and executive function, including in virtual classroom environments where ecologically valid distractors are added.

Control condition

The comparison condition in an experimental study. Used to tell whether the effect seen in the experimental condition is due to the manipulation or to something else (time, attention, expectation).

Cybersickness

Symptoms similar to motion sickness (nausea, disorientation, headache) that some people experience in VR, generally attributed to a sensory mismatch between visual and vestibular cues. Distinct from anxiety symptoms, which it is sometimes confused with. Modern stand-alone headsets have reduced but not eliminated cybersickness; VR session length and content design both matter.

Dose and intensity

Dose is the total amount of therapy delivered (for example, number of sessions or total minutes). Intensity is how much is delivered per session or per week. Studies sometimes find that intensity matters more than total dose for certain outcomes.

Dysarthria

A motor speech difference affecting the muscles involved in speaking. Recurring topic in cognitive-communication after brain injury research.

Dysfluency / disfluency

A research term for moments where the flow of speech changes. In stuttering research the term is often reserved for stuttering-specific moments (what speakers describe as getting stuck or losing control over the speech mechanism), while "non-stuttered disfluencies" refers to the repetitions, revisions, and fillers that appear in everyone's speech.

Ecological validity

The extent to which a research situation resembles the real-world situation it is meant to represent. A virtual cafe intended to generalize to real cafes needs high ecological validity for findings to transfer.

Effect size

A standardized measure of how large the difference between conditions is, independent of sample size. A small effect can still be statistically significant with a large sample; effect size tells you whether the effect is practically meaningful. Common measures include Cohen's d, partial eta squared, and Pearson's r.

Embodied conversational agent (ECA)

A virtual character that perceives and acts in a 3D environment and interacts with the user through spoken dialogue, gesture, and facial expression. The lineage of clinical work on virtual standardized patients descends from this concept.

Experimental study

A study in which researchers actively assign participants to conditions and manipulate variables. Contrasts with observational studies, where researchers only observe.

Fundamental frequency (f0)

The perceived pitch of the voice, measured in Hertz. Used in voice research and in studies of speaking under stress.

Generalization

The transfer of a skill, behavior, or response from one context to another. A goal practiced in a virtual cafe generalizes when the person uses it in a real cafe.

Habituation

The process by which a person's emotional or physiological response to a stimulus decreases with repeated exposure. A central mechanism in exposure-based therapies, and one of the proposed mechanisms by which VR exposure produces lasting change.

Head-mounted display (HMD)

A VR headset worn over the eyes that displays the virtual environment. Meta Quest, HTC Vive, and Valve Index are all HMDs.

Heart rate variability (HRV)

Variation in the time between heartbeats. A physiological measure sometimes used to index stress, arousal, or autonomic regulation.

Immersion

The objective technical properties of a VR system that make it feel involving (visual fidelity, tracking, sound, field of view). Distinct from presence, which is the subjective experience of "being there."

In-virtuo / in-vivo exposure

Two delivery modes of exposure-based therapy. In-virtuo refers to exposure conducted within a virtual environment; in-vivo refers to exposure in the real world. Some research compares the two directly; meta-analyzes and recent non-inferiority trials suggest comparable clinical effects.

Individualized plan

An overarching term for education or therapy plans written for a single student or client. Includes IEP (US), EHCP (UK), OPP (Netherlands), and ILP (Canada, Australia).

Levels of evidence

A hierarchy used in evidence-based practice to rank study designs by how much confidence they warrant. Commonly: Level I (systematic review of RCTs), Level II (RCTs), Level III (non-randomized), Level IV (case series or expert opinion). ASHA uses a closely related system.

Limitations (of a study)

Constraints on what a study's findings can tell us, usually discussed at the end of a paper. Every study has them. Limitations do not mean a study is bad; they mean the conclusions must be weighed against them.

Longitudinal study

A study that follows the same participants over a period of time. Well-suited to questions about change, but more demanding to run than cross-sectional work.

Natural Language Understanding (NLU)

The component of a conversational AI system that maps the user's speech to its intended meaning. The rate-limiting factor for natural virtual-character interaction; in clinical-training research, NLU accuracy below about 75 percent is associated with user frustration, while accuracy at or above about 90 percent supports fluid interaction.

PEDro / PEDro-P scale

A methodological quality scale for appraising RCTs in allied health (PEDro) and speech-language pathology (PEDro-P, 11 items). Used by speechBITE and other evidence databases to rate study rigour.

PICO

A framework for formulating clinical questions: Population, Intervention, Comparison, Outcome. Used to decide what evidence is relevant. Variants include PICOT (adds Time) and PEO (Population, Exposure, Outcome) for qualitative and observational work.

Place illusion / plausibility illusion

The two components of presence proposed by Slater (2009). Place illusion is the sensation of being located in the virtual environment ("I am here"). Plausibility illusion is the sensation that what is happening in the environment is real ("these events are really happening"). Both contribute to whether VR practice produces real-world-like responses.

Plain language summary

A version of a research study written for non-specialist readers. Every entry in this Evidence Hub includes one.

Presence

The subjective feeling of "being there" inside the virtual environment. Distinct from immersion, which is a property of the system rather than the user. Often measured with self-report questionnaires.

Qualitative study

A study that examines experience, meaning, or context using methods like interviews, observation, or thematic analysis, rather than numerical measurement.

Randomized controlled trial (RCT)

A study in which participants are randomly assigned to an intervention or a comparison condition. The random assignment is intended to balance unknown differences between groups.

Real-world transfer

The extent to which gains made in therapy or research appear in everyday life. Closely related to generalization but often measured in terms of participants' own reports of day-to-day situations.

Sample size

The number of participants in a study. Small samples produce less reliable estimates. A finding from a study of 12 people is weaker evidence than the same finding from 120.

Situation (Therapy withVR)

One of 12 virtual environments in Therapy withVR: Café, Classroom, Bakery, Kitchen Table, Break Room, Meeting Room, Speaking Circle, Animal, Room, Auditorium, Reception, and Supermarket. Each situation has defined positions for avatars.

Skin conductance response (SCR)

A physiological measure of autonomic arousal. The skin's conductivity briefly rises when a person is emotionally activated. Used in some VR anxiety studies alongside heart rate.

Social model of disability

A framework that locates the barriers associated with disability in environments and structures, not in the person. Informs how this Evidence Hub and the withVR site talk about communication differences.

Statistical significance

A convention (often p < .05) that says an observed result is unlikely to be due to chance alone. Statistical significance does not automatically mean an effect is large or practically meaningful; see effect size.

Stuttering / stammering

A multidimensional speaking experience in which a person feels stuck or has difficulty moving forward in speech, despite knowing what they want to say. The experience involves more than the observable speech itself; the speaker's own reactions (feelings, thoughts, tension) and listener reactions both shape its day-to-day impact. Increasingly framed as a form of verbal diversity, one of the many ways that people talk, rather than as a deficit. "Stammering" is the term used more often in the UK, Ireland, Australia, and New Zealand; the two refer to the same experience.

Stuttering frequency (%SS)

The percentage of syllables in a speech sample that are stuttered. A common research outcome measure in stuttering studies, though not without limitations.

SUDS (Subjective Units of Distress Scale)

A self-rating scale (typically 0 to 100) on which a person reports their current level of anxiety or distress. Widely used in research on anxiety and exposure.

Systematic review

A structured review of the literature that follows a pre-registered protocol and attempts to find, appraise, and synthesize all relevant studies on a question. Considered a higher level of evidence than any single study.

Threshold Model of Social Influence

A framework proposed by Blascovich and Bailenson predicting when a virtual social interaction will produce real-world-like behavioral effects. The model identifies factors including agency (avatar versus agent), communicative realism, and the level at which the response system is engaged, moderated by the self-relevance of the interaction.

Transformed social interaction

Manipulating self-representation, sensory abilities, or situational context inside VR beyond what real life permits. Examples include being represented by a different-sized avatar, having one's voice altered, or having a virtual audience subtly adjust attentiveness in real time. Used in experimental research on social influence and self-perception.

Validation study

A study designed to show that a tool, measure, or environment does what it is intended to do. Often the first step before a tool can be used confidently in research or practice.

Virtual audience

A group of avatars that represents an audience to the person inside VR. Used in studies of public speaking and in practice for exposure to speaking situations.

Virtual standardized patient (VSP)

A virtual character that simulates a patient or interlocutor for clinical-training purposes. Descended from Howard Barrows's 1963 introduction of the human standardized patient at the University of Southern California. Modern VSPs use embodied conversational agent technology and natural language understanding to support practice of clinical interviewing and difficult conversations.

VR-EBT (virtual reality exposure-based therapy)

Exposure-based therapy delivered in a virtual environment. Used as an umbrella term in the anxiety-disorders literature for any structured exposure approach that uses VR as the medium. Distinguished from broader "VR therapy" by its specific link to exposure as the active mechanism.

Waiting Room (Therapy withVR)

A virtual space the person enters before and after a session inside Therapy withVR. Can display artwork from artists who stutter. Studied implicitly in work on pre-exposure context and transition states.

Within-subjects design

A study design in which each participant experiences every condition. Contrasts with between-subjects. Common in VR validation work because it controls for individual differences.

Gaps and suggestions

If a term you looked for is missing, or if a definition could be clearer, send a note to hello@withvr.app. This glossary is maintained as the hub grows.