The Assessment

The Icosa assessment turns your responses into a complete 20-center personality profile — three tiers of depth, behavioral intelligence captured while you answer, and a structural output rich enough to ground every feature in the system.

One Grid, Three Depths

The Icosa assessment maps your personality across a 4x5 grid: four Capacities (Open, Focus, Bond, Move) by five Domains (Physical, Emotional, Mental, Relational, Spiritual). That grid produces twenty Harmony scores capturing how you process, engage with, and express life. Every assessment, from the fastest to the most comprehensive, generates the same structural output — a complete 20-center profile with coherence scoring, gateway identification, trap detection, and formation classification.

What changes between tiers is precision. A Quick assessment takes a wide-angle snapshot. A Comprehensive assessment uses a telephoto lens. Both capture the same subject — your living personality architecture — but at different levels of resolution.

The confidence number is not a quality judgment. It is a measurement precision indicator that flows through every computation in your profile. A Quick assessment at 0.30 confidence still produces a real, mathematically valid profile — it simply carries wider uncertainty margins. The engine knows what it knows and what it is estimating, and it communicates that distinction clearly.

How Each Tier Works

Quick: The Snapshot

Ten bipolar slider questions, each mapping directly to one or more grid positions. You slide between two poles — I tend to hold back versus I tend to jump in — and the engine converts your position into initial capacity and domain estimates.

Quick is designed for speed without sacrificing structural validity. The profile you get back is a real Icosa profile with all the same constructs (coherence, gateways, traps, formations) computed from genuinely measured data. It is the right choice when you want a fast orientation, a screening check, or a first look before committing to a deeper assessment.

Because each of the ten questions maps to a capacity-level measurement, some grid positions receive overlapping coverage while others are estimated from structural relationships between neighboring centers. The engine is transparent about which positions carry direct measurement and which carry inferred values.

Standard: The Baseline

Thirty-two questions using a split direction-and-magnitude format. First you indicate a direction (which of these two tendencies is more like you?), then you indicate magnitude (how strongly?). This two-step approach captures more nuance than a single slider while keeping the experience focused and fast.

The split format is a deliberate design choice that decouples direction from intensity. Standard is the default for most users. At 0.70 confidence, it provides a solid baseline that is precise enough for meaningful trend tracking across retakes. If your coherence moves from 58 to 67 between two Standard assessments, that shift reflects real structural change — the measurement is sensitive enough to detect it.

Comprehensive: The Full Picture

Ninety-one questions across three question types, each serving a different measurement purpose.

7-point Likert scales capture fine-grained intensity with seven gradations, giving the engine a high-resolution read on absolute magnitude at each grid position.

Differential items present multiple-choice scenarios that disambiguate similar-looking patterns. When two grid positions could explain the same surface behavior, a differential item forces a choice that reveals which position is actually active.

Seed/need items identify both where a pattern originates and which direction it is moving — crucial for distinguishing between someone who is naturally Open and someone who is learning to be. A settled trait and a developmental edge look different in the data, and seed/need items capture that difference.

The combination of these three types at 91 questions produces a profile with full confidence across all 20 grid positions, with enough redundancy to detect internal inconsistency and enough variety to prevent response-set bias.

At 1.00 confidence, Comprehensive produces the highest-resolution profile the system can generate. Every computation runs at full precision. Every behavioral signal is captured at maximum fidelity. This is the tier clinicians use for intake assessments, the tier researchers use for study protocols, and the tier anyone uses when they want the most complete picture of their personality architecture.

Who Is Answering?

The same assessment can be taken from four different perspectives, and the system treats each one as a distinct data source.

Self is you answering about yourself — the default perspective. Your self-perception is valuable because only you have access to your inner experience, but it also carries the blind spots that come with proximity.

Other/Partner is someone answering about you. A partner, close friend, family member, or colleague sees your behavior from the outside. They miss your internal experience but catch patterns you cannot see from inside them.

Clinician is a trained professional observing you through clinical vocabulary. The assessment reframes questions in the third person with clinical precision, and the resulting profile carries the weight of professional observation.

Consensus is not a fifth perspective but a computed aggregate. When multiple reporters have assessed the same person, the system generates a weighted composite using weighted multi-perspective integration — combining self-report, other-report, and clinical observation into a single profile that leverages each source’s strengths while compensating for each source’s limitations.

What Happens While You Answer

The assessment does not simply record your answers. It observes how you answer. Twenty-three behavioral signals across five categories are detected during the assessment process itself.

Timing signals track response latency, hesitation patterns, and pacing. A question you answer in 800 milliseconds tells the system something different from one you deliberate on for twelve seconds — not about correctness, but about certainty and familiarity.

Trajectory signals detect patterns of change across the assessment. Are your responses becoming more extreme as you go? More moderate? More consistent? These trajectories reveal engagement quality and response style.

Sequential signals analyze the relationship between consecutive answers. Sudden reversals, consistent streaks, and oscillating patterns each carry distinct meaning about how you are processing the questions.

These behavioral signals do not replace your answers. They overlay additional information that refines the profile, particularly in areas where your stated responses and your response behavior diverge. A person who rates themselves as calm (stated) but hesitates on every emotion-related question (behavioral) generates a richer, more accurate profile than either data source alone.

How Behavioral Signals Affect Your Profile

Behavioral signals operate as overlays, not overrides. Your direct answers always form the foundation of your profile. Behavioral data adjusts the confidence and interpretation of specific grid positions based on observable patterns.

The amount of behavioral adjustment is tier-dependent. Quick assessments at 0.30 confidence produce fewer behavioral signals because there are fewer questions to observe patterns across. Comprehensive assessments at 1.00 confidence generate the richest behavioral data because 91 questions provide ample space for patterns to emerge.

Save, Resume, and Return

Life interrupts. The assessment saves your progress automatically as you go, with debounced auto-save to both your browser and the server. Close your laptop mid-question, and when you return your answers are exactly where you left them.

This is not just convenience — it preserves behavioral data integrity. The system distinguishes between a natural pause (you walked away to answer the door) and a genuine hesitation (you could not decide between two options). Interrupted sessions resume cleanly without contaminating the behavioral signal layer.

Free to Start

The full assessment is available on the free tier. There is no paywall between you and your first profile. Every tier — Quick, Standard, Comprehensive — is accessible without a subscription.

What requires a subscription is the deeper analysis: AI-generated narratives, longitudinal timeline tracking, centering plans, dyadic compatibility assessments, and framework comparisons. But the core assessment, the computation of your 20-center grid profile with coherence scoring and formation classification, is free. You see your full structural picture before deciding whether you want the interpretive layers built on top of it.

What You Get Back

Regardless of which tier you choose, the assessment produces a complete structural profile: 20 Harmony scores, capacity and domain summaries, a coherence score with band classification, gateway identification, trap detection, basin detection, fault line analysis, pattern matching, dynamics computation, and formation classification. Over 300 computed metrics from a single sitting.

Every question in the assessment is structurally anchored to the 4x5 grid architecture — not selected from a generic item pool. The three tiers do not measure different things. They measure the same twenty centers with increasing resolution, and every piece of your structural profile is grounded in data you provided. The result is a personality portrait that is yours from the first question to the last metric.