A diagnosis system led by the Original 4 Axes, combined with Big Five, RIASEC, and 5 relationship classes
The current public diagnosis lineup uses SOUL/ALLOY's original 4 axes as the main layer, with Big Five and RIASEC as purpose-specific reference assessments. Enneagram is not a standalone public diagnosis entry; it is a supplementary layer for reading inner drives on result and detail pages.
Original 4 Axes
Main diagnosis
12Q / 32Q diagnosis that matches you to the closest material persona. This is the primary public entry.
Big Five
Trait reference
Quantifies personality tendencies across the OCEAN factors and supports compatibility and deeper interpretation.
RIASEC
Career interest assessment
Uses Holland Code to read interests and work-style tendencies. This is a vocational-interest layer, not a personality layer.
Enneagram
Supplementary layer
Used as supporting information for motivations and fears. It is not a standalone public diagnosis entry.
Note: public entry points and internal supplementary layers are separated so current UX and legacy data compatibility can coexist.
Enneagram-derived information is used in persona explanations, secondary compatibility perspectives, and detail views. It is not presented as a standalone public diagnosis entry.
SOUL / ALLOY's main diagnosis uses four original axes, decoupled from Jungian 16-type classification. Each axis is a continuous value from 0 to 100, and the combination of all four becomes your 'coordinates' as a material.
External Link ↔ Internal Cohesion
Reaching out to react, or condensing inward (reactivity / social pull)
Pattern Forecast ↔ Concrete Observation
Reading potential and frontiers, or observing what concretely exists now
Logic Crystal ↔ Resonant Judgment
Deciding by structure and efficiency, or by connection and harmony
Fluid Exploration ↔ Solidified Design
Trying out change, or settling into form for consistency
The 12-question quick version and 32-question detailed version compute your 4-axis coordinates, then find the nearest material (smallest Euclidean distance) among all personas.
Compatibility between you and others, or between personas, is calculated from Big Five (OCEAN) scores. We combine 'similarity' (how alike) and 'complementarity' (how well you cover each other) into an overall score, then classify the relationship into five types.
The smaller the gap across the five factors, the higher the score. Agreement on Neuroticism (N) is weighted most heavily, since differences in emotional stability are the leading cause of relationship wear.
For Extraversion (E) and Openness (O), being 'moderately different' scores higher. Differences that cover each other — introvert × extrovert, broadened perspective — are rewarded.
Note: Optimizing the complementarity parameters (Lever A) corrects the tendency for the top match to be skewed toward 'just like yourself.'
Resonance — Ways of feeling and thinking naturally overlap; easy to be together
Complement — You cover each other's missing pieces
Growth — You learn from differences and spur each other on
Friction — Approaches and values tend to clash (understanding is key)
Storm — Emotional stability differs greatly — the most challenging relationship
We offer multiple diagnosis modes and supplementary analyses. Each mode uses different psychological theories, question counts, and levels of academic reliability.
Theories Used: Original 4 Axes (Quick 12)
4 axes × 3 questions for quick typing. A shortened version with fewer questions, sufficient for identifying general tendencies. Best for trying out the diagnosis.
Theories Used: Mini-IPIP-20
Academically validated short form (Donnellan et al., 2006). Efficiently measures 5 personality factors with 4 items each.
Theories Used: IPIP-50
Internationally used Big Five measurement scale. High internal consistency and test-retest reliability, the academic research standard.
Theories Used: O*NET Interest Profiler / Holland Code
Visualizes interests and work-style tendencies across six vocational-interest types. It is not a personality diagnosis.
A 9-type framework for reading motivations, fears, and desires. In SOUL/ALLOY, it is used to deepen persona detail and result interpretation, not as the main public diagnosis layer.
The most reliable personality model in modern psychology. Captures personality as a continuous spectrum across five factors.
Openness
Conscientiousness
Extraversion
Agreeableness
Neuroticism
Like the Enneagram, uses a hybrid method combining 50% cosine similarity and 50% Euclidean distance.
// 5-dimensional vector: [O, C, E, A, N]
user = [O, C, E, A, N]
persona = [O', C', E', A', N']
// Cosine similarity (direction)
cos_sim = (user · persona) / (||user|| × ||persona||)
// Euclidean similarity (absolute values)
euclidean_sim = 1 - √(Σ(u-p)²) / 223.6
// Maximum distance = √(100² × 5) ≈ 223.6
Similarity = (cos_sim × 0.5 + euclidean_sim × 0.5) × 100
A vocational interest model proposed by John L. Holland in the 1950s. It profiles six vocational interest types (Realistic, Investigative, Artistic, Social, Enterprising, Conventional) and represents career tendencies through a 3-letter Holland Code.
Hands-on work, building things, using tools
Analysis, theory, data, hypothesis testing
Creative expression, originality
Helping, teaching, counseling
Leading, persuading, achieving goals
Organizing, accuracy, following procedures
Uses O*NET Interest Profiler Short Form (60 items). Developed by U.S. Department of Labor. License: CC BY-ND 4.0.
Persona matching uses a weighted score of cosine similarity (60%) and Holland Code overlap (40%).
The psychological theories used in this service have different levels of scientific evidence. Please enjoy the results as entertainment while understanding the standing of each theory.
The most widely validated personality model in psychology. Reproducibility and predictive validity confirmed across thousands of studies, with stable factor structure across cultures.
Questions: Extracted from IPIP (International Personality Item Pool). An open database used in numerous peer-reviewed publications.
Goldberg (1990), Costa & McCrae (1992), John et al. (2008)
This service uses Mini-IPIP (20 items). Accuracy is lower than the full version, but sufficient for identifying tendencies.
Highly validated for measuring vocational interests. Adopted as the foundation of the U.S. Department of Labor's O*NET system. Meta-analyses confirm stable 6-factor structure. Note: this measures vocational interests, not personality.
Questions: O*NET Interest Profiler Short Form (60 items). Developed by U.S. DOL (CC BY-ND 4.0).
Holland (1959, 1997), Rounds & Su (2014), Nye et al. (2012)
Persona matching is a custom implementation. It differs from O*NET's official matching logic.
SOUL/ALLOY's original model capturing personality on 4 axes: External Link/Internal Cohesion, Pattern Forecast/Concrete Observation, Logic Crystal/Resonant Judgment, Fluid Exploration/Solidified Design. The typology-based schemes that inspired these axes are widely adopted but criticized in academic research for poor reproducibility and predictive validity. Scores show normal distribution rather than bimodal, and test-retest reliability is low.
Questions: Uses SOUL/ALLOY's original personality 4-axis items, reconstructed with reference to open-source psychological scales.
Pittenger (1993), Grant (2013)
This service uses continuous scores (0-100) rather than binary type classification, partially mitigating reliability concerns.
A model classifying personality into 9 types based on motivations and fears. Popular for self-understanding, but empirical research is limited.
Questions: Uses OEPS (Open Enneagram Profile Scales). An open-source measurement scale.
Riso & Hudson (1999), Hook et al. (2020), Sutton (2012)
In this service it is positioned as a supplementary layer for deepening result interpretation, not as a standalone public diagnosis.
* This service is not an academic psychological test or vocational aptitude assessment. Results are for self-understanding and entertainment. For important decisions, please consult a professional (clinical psychologist, career counselor, etc.).
The most reliable Big Five (5-factor) model open-source test in psychology. Measures Openness (O), Conscientiousness (C), Extraversion (E), Agreeableness (A), and Neuroticism (N).
An open-source test measuring 9 Enneagram types (using 9-item short version).
Information on metal properties, alloy design, and materials science is based on the following academic and industrial resources.
Basic Materials Science Literature
ASM Handbook, Materials Science and Engineering textbooks
Materials Database
MatWeb, ASM Material Data Sheets
A materials science database operated by Lawrence Berkeley National Laboratory. Used for Pourbaix diagram (potential-pH diagram) and crystal structure data.
The following AI assistant tools were used in the development of this project: