Understand digital traces: an interactive experiment

A small experiment. You are in control.

Only after you start does this box save your topic choices in this browser tab’s session storage. On the three additionally linked example pages, we then count visits, explore-button clicks and time in the active window. The box runs no timer. No other browsing history. No server profile.

The experiment lasts at most 30 minutes. Expired data is removed at the next check on a demo page. “Delete experiment data” lets you clear it sooner. There is no background deletion while the page is closed.

The experiment data is in your browser’s session storage for this tab. That is not a cookie and is not automatically sent with server requests. “Delete experiment data” removes only this experiment.

What happened in your experiment?

These values come only from your visits within this exercise. Opening a page again also counts as a visit. Time is a cautious estimate, not measured attention.

A few traces. A cautious guess.

We do not know your name. But someone could guess possible interests from the topics you visited. You might simply have been curious, testing things or looking for someone else. The guess can be wrong.

Consider this: you click Travel because you want to compare an image. The demo guesses an interest in travel. The click was observed; your motive was not measured. That gap between data and interpretation is the main lesson of the experiment.

How can different websites fit together?

Imagine three independent websites. Each includes the same outside service, perhaps for advertising. That service may connect observations if the browser allows it. It is often called a tracker.

Browser protections can make that connection harder or prevent it. The diagram is only a fictional explanation: our three exercise pages belong to the same website and load no outside tracker.

What does Tor help with – and what does it not?

When you visit a regular website through Tor, it sees the Tor exit’s IP address instead of your own connection’s IP. An IP is an address for a connection. Tor removes that clue to your connection from the website’s view.

Logging in still tells the website which account is yours. Cookies or other saved identifiers may recognise a browser. Tor Browser separates website data and limits other identifying features, but Tor does not prevent every kind of tracking.

Even through Tor, this exercise could observe the buttons clicked here if you start it. Tor does not make actions on the page you are visiting invisible.

What if a name is added later?

A single detail is not a profile. In this fictional example, the provider links identifier A7 to several page visits.

Later, the fictional person Hugo Müller signs in with the same provider and gives their name there. The provider could link the earlier identifier to that account.

A large provider may also connect several of its own services through one shared account. Other websites do not automatically get to see names or cookies. A shared embedded service may connect observations only under suitable conditions. An IP alone does not prove identity.

1. Identifier A7
2. Visits are connected
3. Later: Hugo Müller

Technology helps. Your choices matter too.

Tor protects the route and hides your connection’s IP from regular websites. A VPN works differently and is not an equivalent replacement for Tor.

The pages you visit, where you sign in and the details you share also matter. Good technology and thoughtful handling of your own information belong together.