Silence speaks louder than words - and this might also be true for UXR

by Jasper and Dylan

Last week I had an interesting conversation with a Psychologist turned User Researcher, Dylan Dennis. We had a lovely chat on our interests in user research, collaborating with dev and how to cope with business needs. While discussing our methods, we went a bit deeper into his take on synthesis. As a psychologist by training, he brought an amazing point to the table that I need to share. This one really stuck with me.

As UX researchers, we love our transcripts. Words neatly captured, sentences clearly defined, insights carefully extracted. But when was the last time a user's hesitation, a subtle pause, or an unfinished sentence genuinely informed your research findings?

During our conversation, we found ourselves discussing the nuances often lost in traditional user research methods. Dylan shared insights from his background in psychology that profoundly challenged the way we approach user research synthesis. One particular method, Jeffersonian transcription, emerged as pivotal in reshaping how we interpret user interactions.

This article is an invitation to look beyond words alone, to hear the silence, sense the hesitation, and capture the emotional truths that drive human experience.

Why Traditional User Research Can Fall Short

Typical user research interviews and usability tests prioritise content over context. We document carefully what users say, their explicit feedback, and observable actions. But UX fundamentally revolves around human emotions, behaviours, and experiences, which are rarely linear or purely rational.

Traditional synthesis processes tend to flatten the depth of user experiences, prioritising spoken content over the rich subtext carried in silence, intonation, and hesitation. While, paradoxically, we are exactly looking for those deeper feelings in our interviews.

To be or not to be …Traditional transcripts often omit the richness of human speech: intonations, pauses, hesitations, and even silence. The process of writing our conversations down does not always provide us with the adequate tools to recognise emotion, tone of voice, and the deeper meaning of a conversation. We need to be aware that we are, in fact, limiting our capabilities by conforming to the textual format. And let us be honest, who ever really listens back to their interviews? Ironically, these overlooked moments might tell us more about users' genuine emotions than the words they consciously choose.

Consider this: how many times have you transcribed an interview only to realise later that something feels missing? It is not just about what is said; it is also about what is left unsaid.

Expanding our techniques to capture these subtleties is crucial. Emotion and silence are not mere peripheral aspects; they are fundamental data that shape how people genuinely feel and experience products and services.

Silences Are Data, Too

Silence in user research is not merely a gap in conversation; it carries emotional weight. A user hesitating before answering might indicate discomfort, resistance, confusion, or contemplation. These moments are rich with insights that can fundamentally alter our understanding of their experiences.

Yet, standard transcription methods tend to erase these powerful non-verbal cues entirely, stripping away context crucial for truly empathetic design. By overlooking these silences, we risk missing crucial insights that could significantly inform our designs and ultimately impact user satisfaction and adoption.

Dylan's Case Study - Silence That Speaks Volumes

In a qualitative research study on domestic labour in South Africa, which Dylan contributed to through data collection, silence itself emerged as a meaningful data point. Participants, employers discussing domestic workers in their homes, often communicated more through what they did not say than through explicit statements.

For instance, when asked about fairness in household duties, one participant began with a significant pause, followed by vague responses and deflections. Traditional transcription might have summarised this as non-committal. Yet, Jeffersonian transcription highlighted that this silence was loaded with meaning: a subtle but powerful reflection of underlying social tensions, discomfort, and systemic inequality.

This experience illustrated how capturing non-verbal cues can deepen our understanding of human behaviour and emotion, especially when tackling sensitive or complex social dynamics.

Introducing Jeffersonian Transcription

Jeffersonian transcription, drawn from conversation analysis and ethnomethodology, meticulously documents not only spoken words but also:

Pauses and silence
Overlapping speech
Tone, pitch, and emphasis
Speech rhythm and speed

Rooted in psychological research, this transcription method treats all elements of conversation as data points essential for accurately capturing human communication in all its complexity. By documenting these subtleties, we gain access to deeper insights that can significantly impact research outcomes and design decisions.

Why Intonation and Emotion Matter

Humans communicate emotions primarily through non-verbal cues. We raise our pitch to express surprise or anger, pause to convey uncertainty, and trail off to suggest discomfort. These emotional signals are key to understanding user frustrations, desires, and unspoken needs.

If user research aims to capture authentic user experiences, we must recognise and value these emotional cues as critical insights, not mere nuances.

Psychological Foundations of Emotional Nuance and Silence

Understanding emotional nuance in research is rooted in psychological principles:

Social and Emotional Intelligence: Humans communicate significantly through emotional expressions and non-verbal cues. Research from psychologist Paul Ekman highlights how universally we rely on non-verbal signals to convey complex emotions.
Cognitive Dissonance: Psychologist Leon Festinger's cognitive dissonance theory explains how discomfort or hesitation in speech often indicates internal conflict, a valuable indicator of hidden user pain points.
Ethnomethodology: Developed by sociologist Harold Garfinkel, ethnomethodology underscores that social interactions carry hidden structures and meanings in seemingly trivial interactions, pauses, and conversational shifts.

These psychological foundations confirm that emotional subtleties are not minor additions, they're critical to accurately interpreting user experiences.

Re-examining Traditional Synthesis Through Jeffersonian Transcription

Returning to Dylan's case study, traditional synthesis might have summarised the participant's response as indifferent or satisfied because explicit dissatisfaction was not verbally expressed. Yet through Jeffersonian transcription, the prolonged pause, careful avoidance, and tone of discomfort revealed far greater complexity. The participant was not satisfied; they were navigating complex emotions tied to fear, politeness, and societal expectations.

This revelation underscores a crucial point: user research insights often lie between the words, not in them. If our tools ignore emotional nuance, our findings risk becoming incomplete or misleading.

Common Pitfalls When Overlooking Emotional Nuance

Ignoring emotional nuance and silence can result in substantial research blind spots. Common pitfalls include:

Misinterpreting User Satisfaction: A lack of complaints doesn't always equate to satisfaction. Emotional discomfort or hesitation often indicates deeper, unspoken dissatisfaction.
Overlooking Latent Needs: Users frequently struggle to articulate their deeper needs directly. Emotions and pauses often highlight these hidden areas clearly.
Misleading Validation: Relying solely on what users explicitly say can lead teams to false validations or confirmations, guiding product teams toward ineffective solutions.

Avoiding these pitfalls requires vigilance and intentional listening to emotional and non-verbal cues in user sessions.

Why User Research Techniques Need to Evolve

To truly understand users and their experiences, we need to elevate our approach, moving beyond surface-level transcripts towards methods that embrace emotional depth.

Jeffersonian transcription addresses precisely this gap. By documenting emotional nuances explicitly, we elevate our ability to generate truly empathetic insights. I personally think that AI could be a great mover in this way. It is able to recognise the subtle tones in our voice already. It might be one of the avenues we take as a research community, and it would be a step up from our traditional 'flat' transcript data.

Effective Interview Techniques for Capturing Emotion and Silence

To effectively capture emotional nuance, adopt interview techniques that prioritise empathy and patience:

Allow for Silence: Resist the urge to fill gaps in conversation. Let users pause, reflect, and reveal deeper insights naturally.
Ask Open, Emotion-Focused Questions: Frame questions that explore feelings ("How does that make you feel?" or "Can you describe what went through your mind?") rather than purely functional or factual questions.
Observe Non-Verbal Cues: Watch for facial expressions, body language, and changes in voice or tempo. Non-verbal cues often indicate emotional states users are reluctant or unable to verbalise directly.
Prompt Gently on Hesitation: If a user hesitates significantly, gently ask follow-up questions like, "Is there something else you're thinking about?" or "I noticed you paused, would you like to share more?"

Incorporating these techniques actively surfaces richer emotional insights during user sessions.

Practical Guidelines for Capturing Emotional Nuance in User Research

Capturing emotional nuance and silence in user research requires intentional adjustments to your current practices:

Recording Practices: Always record video or audio of user sessions. While textual notes are helpful, having audio/video captures nuances like hesitation, pitch changes, and silence.
Active Listening: Train yourself and your team to actively listen for changes in vocal tone, pauses, and incomplete thoughts. Make notes about these moments during the sessions.

I think that we should consider AI as just another technology, but one that can be used very badly or the right way. We should at least try to understand the basics of AI and the implications it has on our users and society as a whole.