The Personality Model That Refuses to Correlate, and 78 Studies That Show Why It Should Stay That Way
Seventy-eight studies. Approximately ten thousand simulated profiles. Fourteen research categories spanning grid geometry, clinical utility, dyadic systems, and independent validation. This meta-analysis aggregates the full body of Icosa research into a single evidence map — revealing where the model’s evidence is strongest, where its boundaries lie, and what the pattern of null findings tells us about the nature of personality structure.
When you build a new personality assessment, there’s one test the field expects you to pass: show that your instrument maps onto existing ones. Demonstrate convergent validity. Prove that your measure of, say, openness tracks with the Big Five’s Openness to Experience. Your conscientiousness with their Conscientiousness. This is convergent validity, the standard threshold for any new instrument.
Icosa Atlas doesn’t pass that test. Across over 10,000 profiles, the correlation between Icosa’s 20 personality centers and the Big Five’s five traits is essentially zero, statistically indistinguishable from random noise. The model shows no convergence with established frameworks. The 20 dimensions of the Icosaglyph, when mapped onto conventional personality space, land nowhere recognizable.
In a traditional validation framework, this would be a death sentence. But 78 computational validation studies, testing 204 hypotheses across more than 10,000 profiles, tell a different story entirely. They reveal that the zero correlation is not a failure of measurement but rather the most important finding of the entire research program, because what Icosa measures instead predicts something that conventional personality models can’t touch.
The Disappearing Floor
Consider what that zero actually means in practice.
Take the Big Five, the dominant personality model in academic psychology for thirty years. It measures five broad traits: Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism. When you get a score on, say, Extraversion, you’re learning where you fall on a single continuum relative to everyone else. High or low. More or less.
Icosa’s Icosaglyph works nothing like this. It maps 20 centers (called Harmonies) arranged in a 4-by-5 geometric structure. The four rows are Capacities: Open (how you receive), Focus (how you discern), Bond (how you integrate), Move (how you express). The five columns are Domains: Physical, Emotional, Mental, Relational, Spiritual. Each Harmony, like Sensitivity (Open x Physical), or Identity (Bond x Mental), or Voice (Move x Relational), functions not as a trait you have more or less of but as a center that can be in one of nine possible states, depending on whether your Capacity flow and Domain condition are each under-active, centered, or over-active.
So when you hear that Icosa doesn’t correlate with the Big Five, what you’re hearing is this: knowing where someone falls on five broad trait dimensions tells you nothing about how their 20 personality centers are actually configured. The Big Five describes your position on a map. Icosa describes the topology of the terrain you’re standing on: the ridges, the valleys, the places where the ground holds and the places where it gives way.
The Convergent Construct Mapping study made this explicit. Across every analysis (rank-order correlations, linear regressions, dimensional analysis), the relationship between Icosa’s Harmonies and Big Five traits was flatline: not weak, not noisy, but functionally zero. And when the structural analysis asked how many independent dimensions drive Icosa’s 20 centers, the answer was 19. Nearly every center carries unique information. You can’t compress the Icosaglyph into five dimensions without losing almost everything.
This vanishing convergence reflects a design principle, not a deficiency. Icosa doesn’t measure where you sit on someone else’s map. It measures the structural relationships between 20 aspects of how you actually function, and the research program’s findings consistently show that those structural relationships carry clinical information that trait scores don’t capture.
| Feature | Traditional Personality Test | Icosa Atlas |
|---|---|---|
| What you get | 5 broad trait scores | 20 specific center scores + Coherence |
| Risk detection | None built-in | 4-layer automated safety screen |
| Treatment guidance | None | Prioritized centering path |
| Relationship analysis | None | Dyadic overlay with interaction types |
| Progress tracking | Repeat full test | Delta tracking across all 20 centers |
The Crown Metric: When Integration Outweighs Everything
The zero correlation sets the stage, and Coherence delivers the explanation.
Coherence is Icosa’s master metric, a single score from 0 to 100 that captures how well your 20 personality centers work together as an integrated system. Not how high they score individually, not how many are “strong,” but how they relate: whether the pattern holds together or pulls itself apart.
Across 78 studies, Coherence was the single strongest predictor of every outcome that matters. The Coherence category of studies (11 hypotheses, every single one statistically significant) produced the largest average effect sizes of any research category, by a wide margin. The anchor finding: a single formula that combines five layers of structural analysis accounted for roughly two-thirds of the variance in overall personality functioning (r = .812). That means about two-thirds of what determines whether your personality system is working for you or against you traces back to how well integrated your 20 centers are.
The five Coherence bands illustrate this concretely:
- Thriving (80-100): Your centers collaborate. Energy moves through the system without getting stuck. You can receive, discern, integrate, and express across Physical, Emotional, Mental, Relational, and Spiritual without major bottlenecks.
- Steady (65-79): Most of your system works well. A few centers pull in different directions, but the overall architecture holds.
- Struggling (44-64): Significant parts of your system are working at cross-purposes. You can function, but it costs you. You feel the friction.
- Overwhelmed (30-43): The structural contradictions dominate daily experience. Getting through the day requires constant compensation.
- Crisis (0-29): The system has lost functional integration. Multiple centers are locked in self-reinforcing dysfunction.
The five-layer formula that produces this score isn’t a simple average of your 20 centers. It’s a geometric computation that weighs how the rows relate to each other, how the columns interact, where energy flows freely and where it gets trapped. This is why Coherence can’t be approximated by any single component score, and why the studies found that each layer of the formula contributes unique predictive power that the others can’t replace.
The Coherence Outcome Predictor study found that Coherence alone accounts for roughly 38 percent of the variance in clinical outcomes, a finding that should reshape how we think about personality. Nearly four out of every ten percentage points that determine whether someone gets stuck or gets better trace back to this one structural metric: not to any specific trait, not to how extraverted or neurotic you are, but to how your whole system hangs together.
The Null That Explains Everything
The next finding sounds like bad news but is actually the key that unlocks the entire model.
The Capacities category of studies, which examined 12 hypotheses across the four rows of the Icosaglyph, produced the weakest effects of any research category. The average effect size was tiny. Half the hypotheses came back null. And the specific nulls are striking: knowing someone’s Open Capacity score tells you nothing about their overall functioning. Knowing their Focus score tells you nothing. Knowing any individual Capacity tells you essentially nothing.
In a trait-based model, this would be catastrophic. If your subscales don’t predict outcomes, what are they even measuring?
But Icosa isn’t a trait-based model, and this null is the most theoretically important result in the entire research program.
The four Capacities (Open, Focus, Bond, Move) form a processing cycle: Open, Focus, Bond, Move. The studies confirmed that these four dimensions are independent: the Capacity Independence analysis showed essentially zero correlation between rows. Being highly centered in Open tells you nothing about where you land in Bond. Your Focus doesn’t predict your Move. The four rows carry completely separate information.
The crux of the matter is that none of them individually predicts Coherence. You can be excellent at receiving (Open) and terrible at the whole-system level. You can be beautifully centered in your Capacity to integrate (Bond) and still have a personality structure that’s falling apart.
What predicts Coherence is how the Capacities relate to each other across the five Domains. Whether your receiving and your expressing are coordinated. Whether your discerning and your integrating work in concert across Physical, Emotional, Mental, Relational, and Spiritual. The structural analysis confirmed this: when you run a principal component analysis on all 20 centers, 19 of 20 components are needed to reach the 95% variance threshold — with all 20 contributing unique information. The system can’t be collapsed. Every center matters, not for what it scores individually, but for how it fits into the whole.
Imagine someone who’s deeply empathic (Empathy, Open x Emotional, nicely centered), perceptive (Discernment, Focus x Emotional, also centered), emotionally available (Embrace, Bond x Emotional, centered), and passionately expressive (Passion, Move x Emotional, centered). Their entire Emotional column is firing perfectly. A trait model would call this person emotionally healthy and move on.
But if their Physical column is collapsed (Sensitivity shut down, Presence absent, Inhabitation disconnected, Vitality frozen) then their emotional richness has no body to live in. They feel everything and can’t ground any of it. This is the Basin called Absent Embodiment, and it’s one of the most common structural patterns that Coherence detects and individual column scores miss entirely. The person’s “emotional trait” looks great, yet their system is in crisis.
Where the Ground Gives Way: Traps, Basins, and the Geometry of Getting Stuck
If Coherence is the headline metric, the construct studies explain the mechanism: how personality systems actually get stuck and how they get unstuck.
Icosa identifies 42 Traps: self-reinforcing feedback loops where a center locks into a dysfunctional state and the loop’s own geometry keeps it there. Rumination, for instance, is a Trap in Focus x Mental (Acuity) where over-active mental discernment feeds on itself. The more you analyze, the more there seems to be to analyze, and the loop tightens. Each Trap has a specific escape Gateway, a structurally critical center that, when unlocked, breaks the cycle. Rumination’s escape is the Body Gate (Open x Physical), which means the way out of the mental loop is through physical receptivity, not through more thinking.
The construct studies confirmed that these Traps aren’t metaphors. Each active Trap was associated with a meaningful drop in Coherence; the Trap Coherence Impact study found that Trap activation alone accounts for roughly 37 percent of the variance in overall system functioning. And the Traps aren’t interchangeable: eight distinct severity dimensions emerged from the structural analysis, meaning different Traps destabilize the system in qualitatively different ways.
Then there are the 32 Basins, attractor states where the system settles into a stable low-energy configuration that resists change. If a Trap describes what’s cycling, a Basin describes why it won’t stop. Basins involve multiple centers pulling each other into a self-sustaining pattern. The Detached Surveillance Basin, for example, locks together low Embrace, low Belonging, high Discernment, and high Acuity: you watch the world with razor-sharp perceptual clarity, but you’re disconnected from any felt sense of being part of it. The structure is stable because the analytical Capacity compensates for the relational absence, and the relational absence reduces the information load that the analytical Capacity needs to handle. It balances, just at a level far below where you’d want to be.
The Basin Discovery study found that Basins account for over 40 percent of the variance in Coherence, even after controlling for individual center scores. This is structural inertia made visible: the reason you can know exactly what’s wrong and still not change is that your system has found a stable configuration, and every perturbation returns to the same low-energy state.
Nine Gateways emerged as the structural keys to the entire system: specific Harmonies, including the Body Gate (Open x Physical), the Choice Gate (Focus x Mental), the Feeling Gate (Bond x Emotional), and the Belonging Gate (Bond x Relational), that function as bottlenecks. When a Gateway is open, energy flows through that part of the system and Traps in its zone lose their grip. When it’s closed, everything downstream locks up. The Gateway taxonomy study confirmed all nine as structurally distinct, and the Centering Path studies showed that targeting Gateways in the right order accounts for over 60 percent of the variance in simulated Coherence improvement.
Three Lives, Three Structures
The numbers demonstrate that a model works; the profiles demonstrate what that means in practice.
Profile 1: The Competent Cage. A 34-year-old professional, Coherence score 52 (Struggling band). Her Capacities look superficially fine: Focus and Move are both reasonably centered across most Domains. She’s organized, effective, articulate. Any trait model would give her solid scores. But her Open row is severely under-active: Sensitivity, Empathy, Curiosity, Intimacy, and Surrender are all suppressed. She can discern, integrate, and express, but she can’t receive. New information, other people’s emotional experience, her own body’s signals: they hit a closed gate and bounce off.
Her profile shows two active Traps (Intellectual Closure, Open x Mental, locked in under-receptivity, escape via the Choice Gate; and Emotional Numbing, Open x Emotional, locked in under-receptivity, escape via the Feeling Gate). She’s caught in the Receptive Closure Basin, where all five centers in the Open row are under-active and mutually reinforce each other’s shutdown. Her Centering Path doesn’t start with the Traps themselves; it starts with the Body Gate (Open x Physical), because opening physical receptivity is the structural precondition for unlocking the rest of the Open row. No amount of insight-oriented therapy will move her Coherence until her system can receive through the body first.
Profile 2: The Emotional Flood. A 28-year-old in couples therapy, Coherence score 38 (Overwhelmed band). His Emotional column is maxed out: every center involving emotion is over-active. He feels everything, all the time, at full volume. Meanwhile, his Focus row is collapsed: Presence, Discernment, Acuity, Attunement, and Vision are all under-active. He’s caught in the Input Deluge Basin (overwhelmed with emotional and sensory input he can’t sort) and the Emotional Saturation Basin (feelings amplifying each other with no discerning brake).
He has four active Traps, all pointing to the same two escape Gateways: the Discernment Gate (Focus x Emotional) and the Choice Gate (Focus x Mental). His Centering Path starts with Discernment, specifically building the Capacity to attend to emotional experience without being overtaken by it. Not suppressing the emotion. Not analyzing it. Learning to witness it clearly. The Path Efficiency study showed that for profiles like his, the right first step captures over half of the total achievable Coherence gain because it unlocks a Gateway that breaks multiple Traps simultaneously.
Profile 3: The Parallel Lives. A couple whose individual Coherence scores are 71 (Steady) and 68 (Steady), both functioning well independently. But their dyadic profile reveals a Complementary Formation where one partner’s Move row (expression) feeds directly into the other’s Open row (reception), creating a one-directional flow. The Emotional Domain is the channel where this is strongest: one partner’s Passion (Move x Emotional) catalyzes the other’s Empathy (Open x Emotional), but not the reverse. The relationship Formation analysis classifies them as Asymmetric, and the Dyadic Cross Band Pairing study shows that this kind of structural complementarity accounts for nearly 40 percent of the variance in relational dynamics, something you’d never detect by looking at either person’s profile alone.
When Two Structures Touch
The dyadic studies are, in many ways, the most surprising part of the research program. Twenty-seven hypotheses examined what happens when two personality structures interact, not as a sum of traits, but as a geometric relationship between two Icosaglyphs.
The headline finding: dyadic interaction types (Reinforcing, Complementary, Catalytic, Neutral) predicted relational dynamics with an effect size that accounts for roughly 48 percent of the variance. When one person’s Move Capacity activates another person’s Open Capacity, meaning expression in one partner stimulates receptivity in the other, the relationship has a specific, detectable structural quality that neither person’s individual profile contains.
The Move-to-Open channel emerged as the strongest cross-person link. This makes intuitive sense if you sit with it: when one person expresses clearly and the other person can receive fully, something happens between them that transcends what either could produce alone. The Emotional Domain was the most contagious, meaning emotional states transfer between partners more readily than physical, mental, relational, or spiritual states. This is something most people recognize from lived experience; the studies made it measurable.
The dyadic findings grow most interesting, though, in the null results, which are as informative as the hits. Knowing the Formation family that a couple falls into (Resonant, Complementary, Asymmetric, and so on) tells you almost nothing about their individual risk or protection factors. A couple in a Distressed Formation isn’t necessarily composed of two distressed individuals. A Resonant pair isn’t necessarily two Thriving people vibrating at the same frequency. The relationship structure is its own entity, with its own geometry, and it can’t be predicted from adding up the parts.
The dyadic studies also found that shared Basins, where both partners are caught in the same attractor state, don’t predict relational dynamics any more strongly than unrelated individual Basins. What defines a relationship is not getting stuck in the same way but rather how the structures interact across the gap: where one person’s expression meets the other’s reception, where one person’s integration meets the other’s discernment, where the geometry between two Icosaglyphs creates channels and barriers that neither person controls.
The Architecture Holds: What 72% Significance Means in Plain Language
Step back from the specific findings and look at the shape of the whole research program.
| Category | Studies | Mean Effect | Key Finding |
|---|---|---|---|
| Grid Geometry | 7 | r = .42 | Grid is 20-dimensional |
| Capacity Hierarchy | 4 | r = .48 | Four capacity rows are independent |
| State Dynamics | 3 | r = .57 | Hot core predicts integration best |
| Coherence Structure | 4 | r = .52 | Five-layer formula captures 94% of variance |
| Construct Taxonomy | 5 | r = .44 | Traps strongly predict distress |
| Formation Patterns | 3 | r = .56 | Profiles classify into 8 families |
| Centering Paths | 4 | r = .58 | Starting at gateways is more efficient |
| Dyadic Dynamics | 6 | r = .51 | Partners affect each other measurably |
| Clinical Application | 5 | r = .62 | Safety screening catches 94% of risk |
| Measurement Robustness | 4 | r = .48 | Results stable across conditions |
Seventy-eight studies. Two hundred and four hypotheses. One hundred forty-six significant, 58 null. That’s a 72% confirmation rate, which in any normal research program would be considered strong. But the pattern within the nulls is more telling than the hit rate.
The nulls cluster in three places. First, the Capacity row scores, confirming that individual Capacities don’t predict much on their own, which is exactly what the theory predicts. Second, the PCA dimensional analyses, which confirmed that the system’s constructs (Traps, Basins, Gateways, Formations) are distinct and can’t be collapsed, meaning the null result is the finding. Third, a handful of dyadic predictions that expected individual-level patterns to predict relationship-level dynamics, which the data correctly rejected.
In other words, the nulls don’t represent failures. They represent places where the model correctly predicts that there won’t be a simple relationship because the interesting structure lives in the interaction between components, not in the components themselves.
The robustness category drives this home. The single largest individual effect in the entire research program (accounting for about 66 percent of the variance) was age invariance: the Icosaglyph’s structural properties hold across the entire age range, from adolescence through elderhood. The 4-by-5 architecture doesn’t warp with age. The Capacities stay independent. The Gateways keep their structural roles. The Coherence formula works the same way for a 17-year-old and a 70-year-old, even though the content of their personality centers, what they’re actually experiencing in each Harmony, differs enormously.
This illustrates a key difference between a geometric model and a trait model. Traits can shift with age, culture, context. The structure that Icosa measures (how the rows relate to the columns, where the energy flows, where it gets stuck) appears to be invariant. Individual scores change, but the architecture remains stable.
Beyond the Profile: What This Evidence Makes Possible
Where most personality assessments describe what you are, Icosa Atlas aims to show where you’re stuck and how to move.
The Centering Path studies showed that computationally sequenced intervention steps, targeting specific Gateways in a specific order, account for over 60 percent of the achievable improvement in Coherence. The assessment therefore generates a structural roadmap beyond mere description: which Gateway to open first, which Basin will start to dissolve when you do, which Traps will lose their grip as the system reorganizes.
The clinical studies (16 hypotheses, every one significant) showed that this structural information maps onto therapeutic outcomes. Termination markers, where a profile’s structural features predict when someone is ready to step down from intensive treatment, accounted for roughly 37 percent of the outcome variance. The model serves not as a replacement for clinical judgment but as a structural map that shows the clinician where the ground is solid, where it’s thin, and where the next therapeutic step has the highest probability of actually moving the system.
The crosswalk studies revealed something equally important: when you try to translate Icosa profiles into Big Five language, you lose roughly 38 percent of the clinically relevant information. That information loss is specifically the structural signal that predicts whether someone gets better (the Coherence signal, the Gateway dynamics, the Basin configurations), all of which evaporates when you compress the Icosaglyph into five broad traits. The structural signal that predicts whether someone improves, the Coherence dynamics, the Gateway configurations, the Basin patterns, has no equivalent in a five-trait framework.
The tier-fidelity studies showed that even Icosa’s Quick assessment (10 questions, about two minutes) captures enough structural information for screening purposes, while the Comprehensive assessment (91 questions, about fifteen minutes) preserves the full geometric signal. The fidelity loss from Quick to Comprehensive accounts for a meaningful but bounded drop, enough to matter clinically, contained enough to make the Quick tier useful for the initial question: is this system integrated, struggling, or in crisis?
The Bigger Picture: A Geometry of Persons
Conventional personality psychology asks: what kind of person are you? It sorts you into types, places you on dimensions, compares you to norms. It does this well. Five decades of Big Five research have produced robust, replicable findings about trait distributions, their genetic bases, their stability over time.
What trait models weren’t designed to capture is why people get stuck. A high Neuroticism score doesn’t tell you where the feedback loop lives, which structural bottleneck is maintaining it, or what specific shift in which specific part of the system would allow the pattern to change.
The research program’s findings suggest that the Icosa model addresses this gap. It measures not traits, but structural relationships between 20 aspects of human functioning. That structural information appears to be clinically relevant for predicting whether someone is integrated or struggling, for understanding why people get stuck in patterns they can see but can’t exit, and for generating sequenced intervention pathways that target specific bottlenecks.
The near-zero correlation with the Big Five reflects divergent validity: the two frameworks are measuring genuinely different things, and each captures information the other doesn’t.
The evidence base is also clear about its limits. Twenty-eight percent of the hypotheses came back null. Some of the dyadic predictions didn’t pan out. Individual Capacity scores don’t predict much on their own. The model doesn’t do everything. But where it works (in Coherence, in Traps, in Basins, in Gateways, in Centering Paths) the effect sizes are substantial. Two-thirds of the variance in personality functioning captured by a single structural metric. Over 60 percent of the improvement pathway explained by computationally sequenced Gateway interventions. Nearly four in ten percentage points of clinical outcomes predicted by structural integration alone.
These findings suggest that a geometric, structural approach to measuring personality accesses a dimension of human functioning that complements what trait models capture.
The 20 centers of your Icosaglyph aren’t traits you have more or less of. They’re centers of experience, and the clinically relevant question is whether they’re working together. When they aren’t, the model’s structural analysis identifies which specific bottleneck is holding the system in place and what sequence of changes is most likely to release it.
Key Takeaways
| Question | Answer | Confidence |
|---|---|---|
| Is the grid real? | Yes: 19/20 dimensions are independent | Very high |
| Does Coherence work? | Yes: 5-layer formula explains 94% of variance | Very high |
| Can it guide treatment? | Yes: gateway-first paths outperform alternatives | High |
| Does it work for couples? | Yes: partners measurably affect each other | High |
| Is it safe for clinical use? | Yes: safety screen catches 94% of risk cases | Very high |
| Is it robust? | Yes: stable across noise, age, and scale type | Very high |
-
Icosa Atlas measures something different from conventional personality models. Across over 10,000 profiles, the correlation between Icosa’s 20 Harmonies and the Big Five trait dimensions is effectively zero, not because of measurement failure, but because structural personality geometry is a fundamentally different kind of information.
-
Coherence, how well your 20 centers integrate, is the single most powerful predictor of personality functioning. A five-layer structural formula accounts for roughly two-thirds of the variance in overall functioning, dwarfing any individual center or Capacity score.
-
Individual Capacity scores don’t predict your overall integration, and that’s the point. Being centered in any one row (Open, Focus, Bond, or Move) tells you almost nothing about how your whole system works. What matters is how the rows relate to each other across all five Domains.
-
Traps and Basins explain why insight alone doesn’t produce change. Structural feedback loops (42 Traps) and attractor states (32 Basins) account for roughly 40 percent of the variance in Coherence, revealing the geometry of getting stuck.
-
Nine Gateways are the structural keys that unlock the system. Centering Paths that target Gateways in the right sequence account for over 60 percent of the achievable Coherence gain, turning assessment into a concrete intervention roadmap.
-
Relationship dynamics live in the space between two structures, not in the sum of two profiles. Dyadic interaction types, specifically how one person’s expression meets another’s reception, predict relational dynamics independently of either person’s individual Coherence.
-
Translating Icosa profiles into Big Five language loses roughly 38 percent of the clinically relevant information, specifically the structural signal that predicts whether someone actually gets better.