Log in or create a free Rosenverse account to watch this video.
Log in Create free account100s of community videos are available to free members. Conference talks are generally available to Gold members.
What co-speech gestures reveal about users’ thinking during interviews
Summary
Key Insights
-
•
Gestures provide a non-verbal window into users' mental models, often revealing thoughts not expressed in speech.
-
•
Representational gestures, especially those made in personal gesture space, indicate cognitive processes and implicit imagery.
-
•
Users commonly hold multiple, context-dependent mental models rather than a single static one.
-
•
In interviews, interviewer gestures increase participant gesturing and improve conversational rapport.
-
•
Video interviews pose challenges for capturing gestures fully; positioning and prompting can mitigate this.
-
•
Speech-gesture mismatches often signal ongoing mental model construction or word searching by users.
-
•
Gestures can reveal emotional attachment or disengagement with technology, influencing adoption and retention.
-
•
Mental models can be anchored by recent technology prototypes, such as chat GPT for AI understanding.
-
•
Structured interview protocols that elicit gesturing and separate talking from tool use optimize gesture data collection.
-
•
Open source motion tracking and gesture analysis tools can aid qualitative research by quantifying gesture patterns.
Notable Quotes
"Gestures are a window to the mind."
"Gesture and speech form an integrated system; they reinforce one another."
"We’re really leaving half of our data on the table by not attending to gesture when eliciting mental models."
"Gesture is not computer and smartphone gestures, but spontaneous movements people make with hands and arms."
"Four me gestures happen in that personal gesture space and serve as thinking tools for the speaker."
"When gestures and speech mismatch, it often means the speaker is still refining their mental model."
"Without looking at the gesture, we would have come to a less complete mental model."
"Users have more than one mental model; they can be constructed on the fly depending on context."
"The degree to which the user feels in control with an intelligent agent brings up conversational mental models."
"The more you gesture, the more your interviewee will gesture."
Or choose a question:
More Videos
"You’re constantly training people and optimizing protocols even after successful implementation."
Sofia QuinteroBeyond Tools: The Messy Business of Implementing Research Repositories
March 10, 2022
"Design can no longer sit in a corner and complain about not having a seat at the table; partnerships enable progress."
Sharbani DharBreathing Room for Delight
January 8, 2024
"The opposite of doom scrolling is hope. Hope is active. It’s about claiming our agency in a moment that may be difficult."
James LangHopeful Futures for UX Research
January 30, 2026
"The hardest job I ever had was redesigning an organization, not just a system."
Dana ChisnellThe Sensemaking Business
March 10, 2026
"Our skills matrix became the center of gravity for our practice—it started writing job descriptions and guiding recruitment."
Liam ThurstonWhy Your Design Team Is Quitting, And How To Fix It
June 10, 2022
"At the end of the day, we’re organizing information in the service of humans—it feels new, but it’s always been about designing for people."
Paz PerezContext Architecture for AI systems
September 11, 2026
"You can’t collaborate without communicating upfront and sharing context."
Husani OakleyTheme Two Intro
June 6, 2023
"The more complex the task, the less likely it is to be successful with current AI."
Sarah BarrettAI in Real Life: Using LLMs to Turbocharge Microsoft Learn
February 13, 2025
"If any of those adjectives do not apply to how you’re feeling right now, we wanna know."
Louis Rosenfeld Bria AlexanderOpening Remarks
October 2, 2023
Latest Books All books
Dig deeper with the Rosenbot
What is continuity architecture and why is it important across AI-driven journeys?
How does AI integration impact team burnout and workflow complexity in large design organizations?
How can organizations prevent design system knowledge from being an undocumented API surface reliant on specific individuals?