The method, and its limits
The scoring rules are public. The thresholds are design assumptions. Nothing here is a percentile.
What the index is
Each axis is reported on a directional index from 0 to 100. Zero is the strongest modeled inclination toward the left pole. One hundred is the strongest modeled inclination toward the right pole. Fifty is the midpoint of the directional answers that were actually given. It is not a measure of neutrality, maturity, or uncertainty.
What is left out of the number
A skipped question is unanswered. “I cannot determine” is a recorded response, counted separately, and it does not move the index. Neither one is entered as 50. Some answers are tagged as alternative positions, such as compatibilism or a nihilist reading of meaning. Those answers are also kept out of the index, because forcing them onto a pole would misdescribe them.
How an answer moves an axis
Directional options carry integer weights on one or more axes. A positive weight pulls toward that axis’s right pole; a negative weight pulls toward the left pole. For the questions you answered directionally, the index rescales the sum of chosen weights against the most left-leaning and most right-leaning options on those same questions. Unanswered and non-directional responses do not sit in the denominator as if they were midpoints. The displayed index is that rescaled value, rounded to the nearest integer.
Coverage, disagreement, and classification
Coverage is the share of applicable questions that received a directional answer. Disagreement is a descriptive ratio: one minus the absolute signed weight divided by the total absolute weight. It is not a probability and it is not a confidence percentage. Provisional thresholds, which are design assumptions, are: below 40, left pole; 40 to 60, mixed; above 60, right pole. If disagreement is at least 0.5 and the index lies strictly between 25 and 75, the axis is treated as mixed even when the index would otherwise fall outside 40 to 60. An axis is insufficient when there is no directional score, when fewer than three directional answers were given, or when coverage is under one half.
From six indices to a code
The six-letter code is a separate step from the numerical profile. It is assigned only when every axis has enough coverage to score. If every axis then falls clearly left or right, the code is definitive. If one or more axes are mixed but each still has a nearer pole, the code is provisional and is shown with an uncertainty warning. If any axis lacks a score, or sits exactly at 50 so that there is no nearer pole, NOEMA withholds the 64-type classification. The eight families are an editorial map of ontology, meaning, and social priority. They are not discovered personality clusters.
How this model could later be examined
Nothing in the current release estimates reliability, validity, or population norms. A later study could preregister the item bank, recruit a sample that is not only self-selected website visitors, and report item difficulty and discrimination, test-retest stability, translation equivalence, and whether the six proposed axes are recoverable as dimensions. Cognitive interviews could test whether respondents understand each item as the editors intend. Differential item functioning could compare the English and Chinese wordings. Thresholds at 40 and 60 should be replaced only by cutoffs justified by that evidence. Until such a dataset exists, NOEMA does not report percentile ranks or confidence percentages. The present instrument is an exploratory philosophical questionnaire.