Technical note v0.1
This page shows only methods, boundaries, and data status designated safe for public display by the assessment’s technical note.
Enneagram Personality Test (Nine Types)
Only public-safe sections provided by the backend are shown. Missing sections do not produce placeholder clutter.
test_goal
The FermatMind Enneagram test offers hypotheses about motivational structure that you can read, observe, and provide feedback on. It first shows the primary candidate, top three candidates, full nine-type profile, and close-candidate comparison, then brings the reading back to real situations. This technical note explains how response signals and interpretation rules generate these modules, and which parts remain theoretical interpretations or product hypotheses.
e105_fc144_forms
The E105 five-point rating form uses intensity ratings to build a full nine-type profile and top three candidates. The FC144 two-option forced-choice form records relative trade-off signals through paired choices. The forms have the same evidence level, but their score sources and the questions they address differ. FC144 does not override E105's primary candidate.
score_space_boundary
E105 scores come from five-point intensity ratings; FC144 scores come from relative preferences in paired choices. They can use the same result structure, but they do not support the same numerical comparison.
dominance_gap
The leading-candidate gap describes the relative distance between the first and second candidates within the same form and score space. It guides the tone used for a clear pattern, close-candidate comparison, or a diffuse distribution. E105 and FC144 use different scoring units, so this gap is not a common yardstick across forms, populations, or time. It is also not a measurement-error range, statistical confidence interval, norm-based rank, proof of external validity, or certainty about personality.
confidence_band
Interpretation stability controls the result page's tone, module visibility, and suggested actions. A clear pattern means the interpretation is relatively concentrated within this form on this occasion; close-candidate comparison calls for reading candidates side by side; a diffuse distribution calls for less certainty about a single number; and a response-quality boundary means the signals do not support a strong interpretation.
close_call
When the first and second candidates are close, the page offers a side-by-side comparison and a 7-day observation task, rather than turning a small gap into a definite conclusion.
diffuse
A diffuse distribution means several type signals appear together, and the first candidate and neighboring candidates do not yet support a strong primary-candidate narrative. The page shifts attention to shared themes among the top three candidates, the three-center distribution, recent counterexamples, and observation tasks that can be reviewed.
low_quality
When quality signals are triggered, the page narrows its interpretive tone and suggests retaking the same form when your state is stable, rather than labeling you.
retake_stability
This technical note will continue tracking retest stability, but it does not promise a stability rate until evidence about samples, time intervals, and response quality is sufficient. Analysis must distinguish same-form retests, agreement across forms, short-term changes in state, and user feedback. A difference in one retest cannot be interpreted as personality change.
e105_fc144_agreement
Agreement between E105 and FC144 describes overlap in the first or top three candidates across the two forms. It does not mean their scores can be compared directly. It describes agreement in candidate ranking only and cannot, on its own, establish the Enneagram model's external validity, clinical validity, or ability to predict career outcomes.
resonance_feedback
Users' day-7 resonance feedback and self-confirmed types enter history as self-observation evidence for research and content calibration, without silently overriding the primary candidate. Feedback is an observation layer, not a system reclassification. It can indicate which narratives need less weight, additional counterexamples, or revised wording, but a sense of resonance cannot be equated directly with measurement correctness.
method_boundaries
This test supports understanding personality patterns and self-observation. It is not for clinical diagnosis, hiring screening, treatment advice, health-level judgments, or fixed personality conclusions. The current version retains wings, arrows, subtypes, and developmental stages only as theoretical background or placeholders for future modules, without hard judgments. This technical note distinguishes product rules, operational metrics being collected, planned stability metrics, and theoretical interpretation frameworks; it does not present them with the same degree of certainty.
privacy
Public explanations use aggregated method metrics only. They do not disclose individual answers, private observation text, or feedback records that identify a user. Shared pages do not show private observations by default, and a user's confirmed type is not treated as the system's final type. Feedback, sample content, and technical metrics follow minimum-necessary use, aggregated display, and traceability. Metrics still being planned or collected show definitions and collection directions, without invented figures.
These entries come from the backend’s method boundaries. Cross-form comparisons and theoretical hints are not presented as firm conclusions.
The E105 five-point Likert form uses rating responses to form a full nine-type profile, top three candidates, a leading-candidate gap, and close-call comparison within one scoring context. It is a starting point for self-observation, and its scores are interpreted only within E105.
evidence_level · descriptive
The FC144 two-option forced-choice form records relative motivational signals in difficult trade-offs through paired choices. It can serve as another observation after close-candidate comparison. The forms have the same evidence level. FC144 does not replace E105's primary candidate, top three candidates, or original scoring facts. Read results within FC144's own form boundaries; differences across forms cannot be interpreted as personality change or a judgment that one assessment is better.
evidence_level · descriptive
E105 and FC144 share an Enneagram interpretation framework, but their forms, response tasks, scoring sources, and error structures differ. They do not share a score space. Result pages may reuse structure and safety boundaries; history, sharing, PDF, and retest entry points must retain each form's provenance. The page does not directly compare their numerical values or interpret differences in the top three candidates across forms as changes in individual state, personality development, or evidence that one is more true.
evidence_level · descriptive
Cross-form comparison is disabled. Unless a validated equivalence mapping becomes available, E105 and FC144 remain in separate scoring contexts, and history, sharing, and result-review surfaces do not infer improvement, decline, health status, or measurement precision from their differences.
evidence_level · descriptive
This result is only for understanding personality patterns, self-observation, and reflective dialogue. Current research is insufficient to treat the Enneagram as a fully validated clinical scale. It is not for mental-health judgments, treatment advice, hiring screening, job matching, ability evaluation, or individual risk assessment. High scores, the primary candidate, top three candidates, close-candidate comparison, diffuse distributions, and response-quality notices are interpretation hypotheses that must be checked against real situations.
evidence_level · descriptive
A user's confirmed type records evidence of their own resonance after observation. It is not a system reclassification or external confirmation. It may enter history to mark the user's perspective for later reading. Sharing and result review show it only as a self-observation note, without overriding the primary candidate, top three candidates, scorer output, or scoring facts from the original responses.
evidence_level · descriptive
When response-quality signals are triggered, the result remains readable, but its interpretation boundaries narrow substantially. The page retains only light observation suggestions and avoids turning a single set of responses into a definite type judgment. It first suggests retaking the same form when your state is stable and you have enough time. Switching forms offers another observation only; it cannot bypass quality limits, seek an ideal answer, prove personality change, or compare score increases and decreases.
evidence_level · descriptive
Diffuse means the nine type signals do not yet form a clear leading pattern. The page lowers certainty about any one number and prioritizes the top three candidates, center distribution, recent counterexamples, and a seven-day observation task. Retesting provides another response sample; it does not automatically correct the result to a particular type.
evidence_level · descriptive
A close-call means the first and second candidates are close. The page turns their differences into side-by-side comparison, counterexample checks, and seven-day observation questions without requiring a forced type decision.
evidence_level · descriptive
Definitions, evidence status, sample requirements, and privacy limits are shown here. A definition is not a measured value.
top_gap_distribution
Distribution of the gap between Top1 and Top2 within the same form, used to describe the basis for clear or close-call interpretation.
Sample requirements: Day7 feedback samples are not required.
Privacy: Only aggregate gap distributions are public; individual raw answers are not.
close_call_rate
Proportion of results whose interpretation scope is close-call.
Sample requirements: Long-term tracking requires stable coverage of the questionnaire.
Privacy: Only aggregate proportions are reported.
diffuse_rate
Proportion of results whose interpretation scope is diffuse.
Sample requirements: Long-term tracking requires stable coverage of the questionnaire.
Privacy: Only aggregate proportions are reported.
low_quality_rate
Proportion of results triggering the low-quality boundary.
Sample requirements: Long-term tracking requires stable coverage of the questionnaire.
Privacy: Uses policy outputs without retrieving raw answers.
pair_frequency
Frequency of close-call pairs in results, used to update the pair library and priorities for close-call coverage.
Sample requirements: Long-term tracking requires stable close-call coverage.
Privacy: Only aggregate frequencies at the pair level are reported.
top1_resonance_rate
Proportion of Day7 observations whose final resonance points to Top1.
Sample requirements: Stable Day7 feedback samples are needed before publishing values.
Privacy: Only aggregate observations are counted; individual confirmations are not shown.
top2_resonance_rate
Proportion of Day7 observations whose final resonance points to Top2.
Sample requirements: Stable Day7 feedback samples are needed before publishing values.
Privacy: Only aggregate observations are counted; individual confirmations are not shown.
retake_consistency_index
Definition of Top1/Top3 consistency for the same person retaking the same form.
Sample requirements: Stable same-form retake samples are needed before publishing values.
Privacy: Only anonymous aggregates are used; individual retake histories are not returned.
e105_fc144_agreement
Definition of Top1/Top3 overlap for the same person taking E105 and FC144.
Sample requirements: Paired samples from both forms are needed before publishing values.
Privacy: Agreement does not mean numerical scores can be directly compared across forms.
close_call_conversion
Definition of subsequent FC144 participation or Day7 feedback after a close-call result.
Sample requirements: Stable samples of subsequent close-call activity are needed before publishing values.
Privacy: Only an aggregate activity funnel is examined, not individual reasons for decisions.
misidentification_matrix
Definition of transitions between initial Top1 and the type identified in Day7 self-observation.
Sample requirements: Longer-term Day7 sample accumulation is needed before publishing values.
Privacy: Only aggregate matrices are public; individual confirmations are not.
observation_completion_rate
Proportion of assigned observations reaching Day7 or 100% completion.
Sample requirements: The current observation table supports computation.
Privacy: Only task-completion rates are counted, not feedback text.
day7_return_rate
Proportion of assigned seven-day observation tasks with submitted Day7 feedback.
Sample requirements: The current observation table supports computation.
Privacy: Only task-return rates are counted, not feedback text.
form_switch_rate
Definition of switching from E105 to FC144 or between Enneagram forms.
Sample requirements: Sufficient form-switch samples are needed before publishing values.
Privacy: Only aggregate proportions are public, not individual form-switch histories.
low_quality_close_call_relation
Joint distribution of low-quality and close-call policy outputs in the same set of results.
Sample requirements: The current policy outputs support joint-distribution computation.
Privacy: Only aggregate policy distributions are examined; individual results are not extrapolated into labels.
See which methods are operational, which still need samples, and which claims are not made.
These limits also apply when you interpret results or add observations.
This test supports personality-pattern reflection and is not for clinical diagnosis or treatment advice.
This version makes no claim of external clinical validity, accuracy, or predictive ability.
This test is not for hiring, promotion, termination, or employment screening.
Wing, arrow, subtype, and health level are not hard system conclusions in this version.
E105 and FC144 share one model but use different scoring spaces and are not compared numerically by default.
The type a user recognizes after observation records personal reflection. It never silently overwrites the measured leading candidate.