Methodology
The questions
The question bank contains 224 questions across 15 topics, including 15 "off-axis" ones. Each possible answer gets a score from -2 (left) to +2 (right). This score is never shown during the test.
The arguments of each side are written with the same length and the same care. When a genuine middle position exists, the centre's argument is also presented; this isn't the case for off-axis issues, or when the middle is only a shade of agreement.
No argument is shown during the test: the score measures your spontaneous opinion, not the effect of the arguments. They are all gathered on the results page, together with your answer to each question. If you retake the test, the questions already seen during your visit are left out of the draw (this list stays in the page and disappears when you close it).
The questions were written in French for a Belgian and French audience, then translated. Some issues are specific to Belgium or France: they are only shown to people from those countries.
The draw
Each test randomly draws 2 questions per topic (1 in the express version) and 2 off-axis questions. Questions are filtered according to your country. When two statements from the same topic are drawn, one is worded so that "agree" leans left and the other so that it leans right, to avoid the bias of agreeing by default.
The calculation
Each answer is scaled between -1 and +1 relative to the most left-wing and most right-wing answers of its own question. Your position is the average of these scores, weighted: weight 1 in general, 0.5 for issues whose left/right alignment is debatable. "No opinion" answers and off-axis questions don't count.
| Position | Label |
|---|---|
| -1 to -0.6 | Left |
| -0.6 to -0.2 | Centre-left |
| -0.2 to +0.2 | Centre |
| +0.2 to +0.6 | Centre-right |
| +0.6 to +1 | Right |
The result is indicative if fewer than 80% of the questions were answered.
Country references
The political centre isn't the same everywhere. Your score is compared with each country's point of balance:
| Country | Reference |
|---|---|
| Belgium | +0.12 |
| France | +0.24 |
| United States | +0.27 |
Each reference is the average position of the country's parties, weighted by the votes they won at the last elections. The positions come from experts who rank each party from 0 (left) to 10 (right); the average is brought back onto the test's scale (5 becomes 0). A positive reference means the country's centre of gravity leans to the right of the middle of the scale.
- Belgium: Chapel Hill Expert Survey 2024, 2024 federal elections (average 5.6 out of 10).
- France: Chapel Hill Expert Survey 2024, 2024 legislative elections (average 6.2 out of 10).
- United States: not covered by the Chapel Hill survey. The gap with France and Belgium is measured by the Global Party Survey 2019 (Pippa Norris, Harvard), then added to their values (average 6.3 out of 10).
These references sum up a political landscape in a single number: they give an order of magnitude, not an exact measurement. They are recalculated with each new survey or election.
Your data
Your score is calculated in your browser. No account, no tracking cookie, no IP address and no identifier is stored. The site's security policy (Content-Security-Policy header) prevents your browser from exchanging data with any server other than ours.
What is counted for everyone, anonymously: page views (day, page, country inferred from the connection, device type, referring site) and the number of tests started and completed (format and chosen country). These are simple counters: they contain no answers and can't be used to identify you.
What is only stored with your consent: your answers to the test and your result. Political opinions are sensitive data under the GDPR (Article 9); so you choose, when starting the test, between "Yes, I'll share" and "No thanks". Nothing is preselected, the test is identical either way, and you can change your choice. Answers are stored without an IP address, without an identifier and with only the date, in a database hosted in the European Union. Because they can no longer be linked to you, they can't be retrieved or deleted once sent. They are only used for overall statistics (distribution of answers, average positions) and are neither sold nor passed on.
Your sharing choice is kept in your browser (local storage, no cookie), only so that you're not asked again.
The details (controller, legal bases, retention period, your rights) are on the Legal notice and privacy page.