Summary
IdeoQuiz measures political views on eight axes and compares the result with reference positions for 26 political traditions. This page explains how the questions were written, how scores are calculated, how matches are chosen, how we tested the system, and what its limits are. We publish this so you can judge the result for yourself.
The eight axes
Each axis runs from −100 to +100 between two poles: Equality (+) vs Markets (−); Authority (+) vs Liberty (−); Progress (+) vs Tradition (−); Global (+) vs Nation (−); Ecology (+) vs Growth (−); Secular (+) vs Religious (−); Revolution (+) vs Reform (−); Local (+) vs Central (−). The sign is only a convention; neither pole is "better".
We chose these axes because each captures a disagreement that appears across many countries and that isn't reducible to the others. The economic and authority axes are the classic two used by many political charts. Culture, world, ecology and faith reflect divides that political scientists have documented as increasingly important. Change and scale separate traditions that otherwise look alike, such as anarchists and Marxist-Leninists, or federalists and centralists.
The questions
The full bank contains 203 items in three formats:
- Statements rated on a five-point scale from Strongly disagree to Strongly agree. On every axis, about a quarter to two-fifths of the statements (5–8 of 20–21) are worded towards the opposite pole, which reduces the effect of a general tendency to agree. We plan to move this closer to an even split in the next item revision.
- Scenarios: short, concrete situations with three or four response options, each linked to positions on one or more axes.
- Forced choices between two values or trade-offs, with no neutral option.
The Quick test uses 60 items: 44 statements, 9 scenarios and 7 forced choices, with at least four items on every axis. The Full test uses all 203.
Some statement formats were inspired by open-source political tests released under the MIT licence (8values, 9axes, SapplyValues); we credit them in our sources. All item wording in our bank is our own, and we didn't copy items from proprietary tests.
From answers to axis scores
- Statements: each answer is converted to a value from −2 to +2 (answer minus 3), multiplied by the item's weight on each axis it measures, with reverse-keyed items flipped.
- Scenarios and forced choices: the chosen option adds twice its weight to each axis it is linked to.
- Normalisation: for each axis, we compute the minimum and maximum totals possible given the items you answered, and express your total as a percentage of that range. The axis score is that percentage × 2 − 100, giving −100 to +100.
Skipped questions are simply left out, so they don't push you towards the middle.
From axis scores to ideology matches
Each tradition has a reference position on all eight axes, set from its core commitments as described in its own literature and standard reference works, and documented on its guide page. We calculate a weighted Euclidean distance between your eight scores and each reference position. The economic, authority and culture axes have a weight of 1.5; the others have a weight of 1. The similarity percentage is 100 minus the distance as a percentage of the largest possible distance. Your top three traditions are the three with the highest similarity.
Reference positions are judgements. Reasonable scholars might place a tradition differently, and every tradition contains internal variety. We publish the positions so they can be scrutinised, and we'll revise them if they are shown to be wrong.
Testing the system
We tested whether the scoring can recover a known position. For each tradition we generated 10 simulated respondents whose answers matched that tradition's reference position, then added random noise to imitate inconsistency, and checked whether the system returned the right tradition. The results on an independent test run were:
| Answer noise | Quick: correct #1 | Quick: in top 3 | Full: correct #1 | Full: in top 3 |
|---|---|---|---|---|
| Low (0.7) | 93.5% | 99.2% | 97.7% | 100% |
| Medium (1.0) | 88.1% | 99.2% | 95.8% | 100% |
| High (1.3) | 76.9% | 96.9% | 91.5% | 100% |
What this does and doesn't show. It shows that the scoring is internally consistent and that 60 items are enough to identify the general area of most profiles, while 203 items separate close traditions better. It doesn't show that the test is valid with real people: simulated respondents are generated from our own model, so they can't reveal problems with how real people interpret questions. We plan to publish anonymised, aggregate reliability statistics (such as internal consistency and test–retest agreement) once enough people have consented to research use, and we'll update this page.
Quality checks in your result
- Acquiescence check: if you agreed with most statements on both sides of an axis, your result shows a note, because the score may partly reflect a tendency to agree.
- Uncertainty range: each axis score shows a range estimated by resampling your own answers. Fewer answers mean wider ranges.
- Close matches: if your top matches are within a few points of each other, the result says so.
Limitations
- Political vocabulary differs between countries; "liberal" is the classic example. We describe positions rather than relying on labels, but cultural context still affects how people read questions.
- Eight axes are still a simplification. Real views are more detailed than any score.
- Reference positions summarise traditions with internal disagreement.
- Your answers can vary with mood, recent news and question order.
- The test describes views. It doesn't measure knowledge, intelligence or moral worth, and it isn't voting advice.
The optional AI report
The optional paid report is generated by an AI model, which is given only your scores, your answers and the relevant entries from our knowledge base. It is instructed to remain neutral, never to mention living politicians or current parties, to quote your answers accurately and to present tensions as interesting rather than as errors. Every report is labelled as AI-generated and passes through automated moderation before delivery. AI can still make mistakes; if you find one, please report it and we'll correct or refund.
Sources
Our ideology guides draw on primary texts and standard histories cited on each page. Scoring is our own. We acknowledge the open-source projects 8values, 9axes and SapplyValues (MIT licence) for formats that inspired ours. We're not affiliated with "The Political Compass" or any other test.