Test Methodology

Every quiz covered by this test methodology draws from one bank of 30 original statements: 25 core value statements, five per dimension, plus five US policy statements for the party-comparison pages. The statements were written for this project in plain, neutral language, and each one expresses a single idea — no double-barreled “X and therefore Y” items.

Within each dimension the statements point in both directions: agreeing with “public services should be expanded” moves the economic axis one way, while agreeing with “private provision should be considered first” moves it back. This bidirectional design limits acquiescence bias — the tendency to agree with whatever is asked — which plagues quizzes where every statement pushes the same way. It is one of the small test methodology choices that quietly improves honesty.

The current statement set carries revision 2026-10-r1. When wording changes under the test methodology, the revision string changes with it, and the change is noted on this page. Statements are intentionally country-neutral except the five US policy items, which explicitly concern federal policy.

Answers use a symmetric five-point scale: strongly disagree (−2), disagree (−1), neutral (0), agree (+1), strongly agree (+2). Skip records no value. Neutral counts as an answered statement that moves nothing; skip removes the statement from that dimension’s calculation entirely — the two are deliberately different, and the test methodology depends on that distinction.

For each dimension, the score is computed only from answered statements. Each answer is multiplied by the statement’s polarity (+1 or −1, fixing which end “agree” pushes toward), summed, and mapped to 0–100: score = 50 + 50 × (sum of signed answers) / (2 × number answered), rounded to the nearest integer. Neutral-everywhere therefore scores exactly 50 on every axis, and fully reversed answer sets produce mirror-image scores — the simplest test methodology check anyone can run.

Two safeguards in the test methodology keep the arithmetic honest. A dimension with fewer than three answered statements reports “insufficient data” instead of a number — the test methodology never fabricates a midpoint. And when fewer than 80% of a quiz’s statements are answered, the results view shows a banner inviting you to answer more, while still displaying what can honestly be computed.

Several pages show a left–right position computed as 0.45 × economic + 0.40 × social + 0.15 × (100 − environment), with the civil and global axes excluded. These weights are editorial choices made once and published in the test methodology — economic posture first, pace of social change second, environmental approach a distant third.

The environment term is inverted (100 − environment) because in current usage, support for regulatory and public-investment routes indexes leftward while preference for voluntary and market routes indexes rightward. That association is a feature of present coalitions, not a law; the weighting exists so the number is reproducible, not so it is authoritative.

The projection returns null when any input dimension lacks data. If you change the weights — and the test methodology invites you to recompute with your own — the same honest answers yield different positions. Publishing the formula is what makes the number an illustration you can argue with rather than an oracle you must accept.

The four party pages compare your five US policy answers with reference positions for each party: the Democratic Party’s published issue pages (democrats.org/where-we-stand) and the Republican Party’s 2024 platform (gop.com/platform-2024), accessed 2026-10-16. Editors compressed each party’s stance per issue into one value on the same −2..+2 scale the quiz uses, following the same test methodology rules as user answers.

Proximity is root-mean-square distance across the issues you answered. When the two overall distances differ by less than 0.25 RMS points, the result is reported as “mixed / no clear match” — a statistical tie is reported as a tie. Fewer than three answered policy statements produce “not scored” rather than a guess. These thresholds are part of the test methodology and are tested automatically before every release.

These references are simplifications by design. Both parties contain substantial internal diversity on every issue; the values describe platform positions, not every officeholder or voter. If an issue’s platform position becomes genuinely ambiguous, the correct handling under the test methodology is “not scored” symmetrically for both parties — never an invented value. Source links and revision dates appear with every party result.

These are informal self-assessment quizzes, not validated psychometric instruments. The statements are original drafts; the scales have not been normed against a population sample; the midpoint of 50 describes this questionnaire’s own range, not the political center of any country. The test methodology is transparent precisely so you can see those limits yourself.

No page on this site recommends a party, candidate, or vote. The party pages describe issue distance to published positions and stop there. Nothing about your age, location, or demographics enters any calculation, because none of it is collected in the first place — a privacy guarantee that is structural to the whole test methodology.

Finally, scores describe positions on the specific statements asked. Political views are contextual, coalition-dependent, and evolvable; a five-point scale cannot carry the trade-offs real policy involves. Where the tests are silent, the test methodology keeps them honestly silent.

All scoring under this test methodology happens in your browser. Answers exist only in the memory of the current page view: no account, no database, no local or session storage, no transmission of answers or scores, and no analytics that see individual responses. Reloading any quiz page erases everything — the strongest privacy rule the test methodology can offer.

The downloadable share card is rendered by your own browser from your own results when you click the button, and is labeled “informal self-assessment.” The card contains a summary of dimensions, never your statement-by-statement answers.

Political opinions are special-category data under the GDPR and comparable frameworks. Local-only processing is the strongest available answer within this test methodology: data that is never collected cannot be leaked, subpoenaed, or sold.