Concept

Knowing you might be watched can affect whether people search, speak or share.

Issue

A question you never asked may matter, but holding back doesn’t tell you what caused it, or whether going ahead would be safe.

Solutions

Understand the evidence and the choices you have, then consider one unasked question.

Say you start typing a question, wonder who might see it, and close the page. You still want an answer. Now there are two questions: what you wanted to know, and what made you hold back.

When people hold back because they think someone might be watching, researchers call it a chilling effect. Studying it means looking for something that might have happened but didn’t. A search nobody made leaves less to look at than one they did.

This module is for adults with a fair question they’ve held back from asking, partly because someone might see. Use it only if thinking it through is safe.

What you’ll get from this module

In this module

By the end of this module, you’ll know more about what made you hold back. You’ll also know what choices you really have. We’ll look at Wikipedia traffic, a Swiss experiment, Canadian interviews and monitoring in restaurants. Then you’ll think about what the findings mean, pick a situation from daily life and work through one question. You may reach a clearer choice, or name a limit you can’t get past.

What can researchers learn from what people don’t do?

Observational study: a change in information traffic.

Jonathon W. Penney’s 2016 study looked at monthly views of 48 English Wikipedia pages about terrorism. It covered January 2012 to August 2014. The pages were chosen from a list of terrorism keywords, not picked at random.

After news of government surveillance broke in June 2013, the paper reports 526,614 fewer views a month on average. That before-and-after comparison fits the idea that people avoided sensitive information. But it counts page views. It doesn’t identify people who stopped reading.

Penney also looked at trends and at other topics to compare with. To blame the drop on fear of being watched, past traffic must be a fair guide. It has to show what would have happened with no news. Shifting news interest, or people reading the same thing elsewhere, could affect the result.

Why it matters: The study shows a drop in traffic that might otherwise go unnoticed. It can’t tell each reader’s reason, or whether the drop lasted beyond August 2014. That difference matters for your own hesitation too: a change is something to understand, not its own explanation.

Read the study of surveillance publicity and Wikipedia use.

Randomized experiment: making observation noticeable.

Noemi Festic, Moritz Büchi, Michael Latzer and Céline Odermatt’s 2026 paper offers a different kind of comparison. Their analysis included 827 German-speaking Swiss internet users aged 16 and over, taking part for six weeks in 2023.

Each time, people were shown one of two things at random. One drew attention to being watched online. The other was on the same topic but didn’t. Then the researchers measured how at ease people felt sharing, and how they handled set tasks. For two topics, the material didn’t make people feel more watched as planned. The researchers left those topics out.

When people were reminded they were watched, their comfort with sharing views fell by 0.17 points on a 1–5 scale. People were also less likely to say they felt comfortable searching and then pasting a link to a result. The researchers counted those two together. They did it within set tasks, not across all of people’s browsing.

Random assignment makes it more likely that the material itself changed these measured responses. Nobody was penalised in the experiment, though people could still expect consequences. So the chance of being watched can matter before anyone actually punishes anything.

Why it matters: More reminders over the weeks didn’t always go with lower comfort in every task. The experiment shows that making people aware of being watched can change a response in the moment. It doesn’t measure a lasting loss in everyone’s life.

Read the experiment on observation and legitimate communication.

Interview preprint: concern does not always become avoidance.

Martin P. Robillard, Lihn V. Nguyen, Deeksha Arya and Jin L.C. Guo’s 2026 preprint reports interviews with 35 adults in Canada. They talked about searching for health information. Most didn’t describe a time they actually held back or changed a search.

Six talked about situations they imagined might happen. One described a personal pattern, but spoke about what they might do when embarrassed. Imagining you might avoid a search is different from remembering one you actually avoided.

Why it matters: These interviews challenge the idea that everyone holds back important questions. The group was small, and wasn’t chosen at random. What they said can’t tell us how often people avoid a search. Nor can it rule out changes people didn’t notice. The practical lesson: allow the honest answer that you have no real example.

Read the interviews about tracking and health-information searches.

What can a missing search tell you?

Workplace study: some restraint has benefits.

Lamar Pierce, Daniel C. Snow and Andrew McAfee’s 2015 study looked at theft and sales records from 392 restaurants. Five companies ran them. The companies brought in monitoring technology at different times.

The researchers compared changes around the time it arrived and reported less theft. The technology helped managers spot theft and could change how they responded. So the study can’t separate those changes from staff simply knowing they might be watched. Its finding is about harmful behaviour, not about looking for information you have every right to.

Why it matters: This adds a useful point: less activity isn’t always a loss. What stopped, why it mattered and the conditions around it all affect what it means.

Read the study of restaurant monitoring.

Think through a held-back search

What this means for you

The studies give reason to take worry about being watched seriously, without assuming it explains every silence. Start with a fair question you already cared about, rather than hunting for signs of a hidden cost.

Start with something you actually wanted to know

Name the answer or help you wanted. You can use a broad category and leave sensitive details unwritten. If no real question comes to mind, stop with “No concrete avoidance identified.” The Canadian interviews show why that answer belongs in the exercise.

Never use this reflection to take a risky step in an unsafe home or workplace. If someone might really punish or pressure you, noticing that limit doesn’t remove it.

Separate your concern from your explanation

Go back to the made-up moment when you closed the page. You might know that a record exists but only be guessing who will read it, or what they will do. Keeping those apart helps you describe the worry without treating a guess as a fact.

Then think about other reasons you held back. Time, cost and embarrassment count. So do low priority, doubt it would help, or a choice not to share. Several can be true at once. Picking one explanation too fast would repeat the mistake of treating missing traffic as a known motive.

You can describe your result as “Observation concern was part of this decision,” “Uncertain,” or “No concrete avoidance identified.” These are descriptions, not scores or instructions to undo your choice.

Recognizing a concern does not require action

Noticing a worry doesn’t mean you have to undo your decision. If going ahead would be unsafe, accept that limit without pushing yourself forward.

The restaurant study also reminds us to ask what kind of action is being held back. This module is about fair questions and seeking help. It gives no reason to treat every rule or form of oversight as harmful.

Stop the reflection if it doesn’t lead to a useful decision. You don’t need to share the question or prove you’ve lost something.

What is behind a question you left?

How about your daily life?

Have you put off a question you wanted answered, softened it, or dropped it? Health, money, legal help and family matters can be private prompts, without writing down the details behind them.

List situations only if that feels right for you. One good idea is enough.

Broad situationAnswer or help I wanted
________________________________________
________________________________________

My chosen situation: ____________________

Take one situation into the exercise. Choosing it doesn’t mean you should go ahead with the original step. If none comes to mind, choose “No concrete avoidance identified” and stop.

Consider one unasked question

Use your chosen situation. If you started here, choose one fair question, or one step towards help, that you already wanted to take but held back on. Keep only a broad category; you don’t need any sensitive detail or to send anything.

Never take a risky step in an unsafe home or workplace for this exercise. It can’t show who is watching, measure what you’ve lost, or make things safe or private. Stop if the reflection doesn’t lead to a useful decision.

  1. Name the question and its purpose. What answer or help did you want? If no real question comes to mind, select “No concrete avoidance identified” and stop.
    Broad question and wanted help: ____________________
  2. Describe your concern about observation. Separate what you know about a record or a watcher from what you assume.
    Known information / assumptions: ____________________
  3. Consider other reasons. Include time, cost, priority, embarrassment, not knowing if it would help, or a deliberate choice not to share. Several reasons can sit together.
    Other reasons: ____________________
  4. Choose a description. Select “Observation concern was part of this decision,” “Uncertain,” or “No concrete avoidance identified.” There’s no score, and no duty to undo your choice.
    My description: ____________________
  5. Identify what remains unresolved. Note what you still don’t know about your worry. Noticing it doesn’t make things safe, or mean you have to act.
    If going ahead would be unsafe, write down the limit without taking a risky step. You don’t need to share anything to finish.
    Unanswered question or unresolved constraint: ____________________

Ask Jing

Become a Jingius.

Think through one question you left

Loading interactive tool…

Your next step

You have finished when you have a description of this decision and any open question or limit. A question left unasked is not proof that you lost something.

A note on the evidence

Traffic changes, set tasks and interviews answer different questions. Together, they can’t tell you how many chances you personally have lost. The evidence allows for both holding back because of being watched, and cases where people report no real avoidance.

How we prepared this module

This module brings together studies of searching, sharing and being watched at work. A made-up example of closing a page shows a decision without treating it as observed evidence. The written exercise helps you think through one situation and its real limits.

Sources