<![CDATA[</tr>
<tr><td><strong>2</strong></td><td>Structural</td><td>Data exchanged in defined format/syntax (HL7 FHIR, C-CDA)</td><td style='color:var(--danger)'>"Receiving system has the file but cannot parse fields"</td></tr>
<tr><td><strong>3</strong></td><td>Semantic</td><td>Receiving system interprets data with same meaning (SNOMED CT, LOINC, ICD-10)</td><td style='color:var(--danger)'>"Both store 'HbA1c' but disagree on units and reference ranges"</td></tr>
<tr><td><strong>4</strong></td><td>Organizational</td><td>Governance, policy, trust, consent frameworks across organizations</td><td style='color:var(--danger)'>"Technically perfect exchange, legally blocked"</td></tr>
</tbody>
</table>
</div>
<div class='callout accent'>
<span class='callout-icon'>π</span>
<div class='callout-content'><p><strong>Hierarchy:</strong> Levels are cumulative β each presupposes the prior. Foundational is necessary but insufficient for semantic; semantic is necessary but insufficient for organizational.</p></div>
</div>
`,
init: () => initLevelDiagram()
};
function buildLevelCards(){
const levels = [
{num:1, name:'Foundational', desc:'Establishes inter-connectivity requirements for systems to securely communicate data.', status:'Partial', statusClass:'status-partial', color:'#58a6ff'},
{num:2, name:'Structural', desc:'Defines format, syntax, and organization of data exchange at the data-field level.', status:'Absent', statusClass:'status-absent', color:'#7ee787'},
{num:3, name:'Semantic', desc:'Provides common underlying models and codification for shared understanding and meaning.', status:'Absent', statusClass:'status-absent', color:'#f0883e'},
{num:4, name:'Organizational', desc:'Includes governance, policy, social, legal, and organizational considerations.', status:'Failing', statusClass:'status-failing', color:'#f85149'}
];
return levels.map((l,i) => `
<div class='level-card' data-level='${l.num}' data-detail='${l.num}' style='cursor:pointer'>
<span class='level-status ${l.statusClass}'>${l.status}</span>
<div class='level-num' style='color:${l.color}'>${String(l.num).padStart(2,'0')}</div>
<div class='level-name'>${l.name}</div>
<div class='level-desc'>${l.desc}</div>
</div>
`).join('');
}
const levelDetails = {
1: {name:'Foundational Interoperability', status:'Partially Achieved', color:'var(--warn)',
content:`<p>Electronic data exchange between SSA, state Disability Determination Services (DDS), some EHRs, and some federal courts is technically possible. SSA's Electronic Records Express (ERE), Health Information Exchange (HIE) infrastructure, and TEFCA's QHINs provide underlying connectivity.</p>
<p><strong>Gap:</strong> The "last mile" to the pro se litigant is weak. Claimants submit paper records, paper SSA-827 forms, unstructured PDFs. From the litigant's vantage, the foundational layer feels like a wall of fax machines.</p>
<div class='callout danger'><span class='callout-icon'>β οΈ</span><div class='callout-content'><p><strong>Failure mode:</strong> "Pipes open to the agency's side, closed on the claimant's side."</p></div></div>`},
2: {name:'Structural Interoperability', status:'Largely Absent', color:'var(--danger)',
content:`<p>SSA internal data models (DDS disability folder, eView, AC database) are internally structured but <strong>not exposed</strong> in standard interoperable formats to claimants, representatives, or AI tools. There is no FHIR-equivalent API for "give me a claimant's complete disability record in structured form."</p>
<p><strong>Gap:</strong> A pro se litigant's medical record exported from SSA may be a flat PDF; CE reports are free-text with no field-level structure.</p>
<div class='callout danger'><span class='callout-icon'>β οΈ</span><div class='callout-content'><p><strong>Failure mode:</strong> "Receiving system has the file but cannot parse the fields."</p></div></div>`},
3: {name:'Semantic Interoperability', status:'Critical Failure', color:'var(--danger)',
content:`<p><strong>The central problem.</strong> Disability evidence is <em>meaning</em> β the meaning of a diagnosis for <em>this claimant's</em> specific functional limitations for <em>this type of work</em>. Three semantic failures:</p>
<ol style='padding-left:20px;color:var(--text2);font-size:.85rem;line-height:1.7'>
<li><strong>Diagnostic β functional translation failure</strong>: A DSM-5 or ICD-10 diagnosis is not a work-disability statement.</li>
<li><strong>Medical-record β legal-issue translation failure</strong>: Same record read by physician (clinical frame), CE provider (adjudicatory frame), DDS (regulatory frame), ALJ (legal frame), court (constitutional frame).</li>
<li><strong>Genomic / neuroimaging translation failure</strong>: Polygenic risk scores, hippocampal volumes, executive-function latent factors β none has an SSA semantic profile.</li>
</ol>
<div class='callout danger'><span class='callout-icon'>β οΈ</span><div class='callout-content'><p><strong>Failure mode:</strong> "Both systems store 'impairment' but disagree on what counts, how severe, and what work it precludes."</p></div></div>`},
4: {name:'Organizational Interoperability', status:'Failing', color:'var(--danger)',
content:`<p>Requires governance, policy, trust, and consent frameworks across organizational boundaries. In disability adjudication, this is the procedural and constitutional regime: APA, Social Security Act, Due Process, Rehabilitation Act Β§504, <em>Mathews v. Eldridge</em>, mandamus remedies.</p>
<p><strong>The data:</strong> 30,000 deaths-while-pending in FY2023; 60β65% federal-court remand rates; two-thirds of remanded cases ultimately awarded benefits.</p>
<div class='callout danger'><span class='callout-icon'>β οΈ</span><div class='callout-content'><p><strong>Failure mode:</strong> "Technically meaningful records, semantically interpretable, blocked from producing timely legal effect."</p></div></div>`}
};
function initLevelDiagram(){
const cards = document.querySelectorAll('.level-card');
const detail = document.getElementById('levelDetail');
cards.forEach(card => {
card.addEventListener('click', () => {
cards.forEach(c => c.classList.remove('active'));
card.classList.add('active');
const lv = parseInt(card.dataset.level);
const d = levelDetails[lv];
detail.innerHTML = `
<h3 style='margin-bottom:8px;color:${d.color}'>Level ${lv} β ${d.name}</h3>
<div style='margin-bottom:12px'><span class='tag ${lv===1?'orange':'red'}' style='font-size:.7rem'>${d.status}</span></div>
${d.content}
`;
detail.classList.add('visible');
});
});
// Auto-select first
cards[0]?.click();
}
sections.diagnostic = {
title: 'Diagnostic Analysis',
render: () => `
<h1 class='page-title'>Diagnostic: The Current Adjudicatory Stack</h1>
<p class='lead'>Assessed against the HIMSS four-level model. Data drawn from the Research Portfolio Exhibit, PSDLEF paper, GAO reports, and SSA workload data.</p>
<div class='table-wrap'>
<table>
<thead><tr><th>Level</th><th>HIMSS Health-IT Analogue</th><th>SSA / Pro Se Status</th><th>Principal Failure Mode</th></tr></thead>
<tbody>
<tr><td><strong>1</strong></td><td>Two systems connected</td><td><span class='tag orange' style='font-size:.66rem'>Partial</span> Agency-side yes, claimant-side no</td><td>Paper submissions, fax, PDF</td></tr>
<tr><td><strong>2</strong></td><td>HL7 FHIR / C-CDA</td><td><span class='tag red' style='font-size:.66rem'>Absent</span> No public API, no SSA-specific profile</td><td>Medical records as opaque PDFs</td></tr>
<tr><td><strong>3</strong></td><td>SNOMED CT / LOINC / ICD-10</td><td><span class='tag red' style='font-size:.66rem'>Largely Absent</span> No shared adjudicator-legible vocabulary for functional impairment</td><td>Diagnosis β work-disability</td></tr>
<tr><td><strong>4</strong></td><td>TEFCA, QHINs, governance</td><td><span class='tag red' style='font-size:.66rem'>Failing</span> 30,000 deaths/yr, 60β65% remand</td><td>Rights exist but no timely correction</td></tr>
</tbody>
</table>
</div>
<h2 class='section-heading'>Diagnostic Visualization</h2>
<div class='chart-container'>
<div class='chart-title'>SSA Adjudication System β Four-Level Maturity</div>
<div class='bar-chart' id='diagChart'></div>
</div>
<div class='callout danger'>
<span class='callout-icon'>π΄</span>
<div class='callout-content'><p><strong>Diagnostic conclusion:</strong> The SSA adjudication system is <strong>Level 1/2 with catastrophic Level 3/4 failure</strong>. PSDLEF is in principle positioned to provide Level 1β4 architecture for pro se litigants β but only if its modules are explicitly re-specified in interoperability terms.</p></div>
</div>
`,
init: () => initDiagChart()
};
function initDiagChart(){
const data = [
{label:'L1 Foundational', pct:45, color:'l1', score:'Partial'},
{label:'L2 Structural', pct:20, color:'l2', score:'Absent'},
{label:'L3 Semantic', pct:10, color:'l3', score:'Absent'},
{label:'L4 Organizational', pct:15, color:'l4', score:'Failing'}
];
const container = document.getElementById('diagChart');
if(!container) return;
container.innerHTML = data.map((d,i) => `
<div class='bar-row'>
<div class='bar-label'>${d.label}</div>
<div class='bar-track'>
<div class='bar-fill ${d.color}' style='width:0%;transition-delay:${i*150}ms' data-width='${d.pct}'>${d.score}</div>
</div>
</div>
`).join('');
requestAnimationFrame(() => {
setTimeout(() => {
container.querySelectorAll('.bar-fill').forEach(bar => {
bar.style.width = bar.dataset.width + '%';
});
}, 100);
});
}
sections.sil = {
title: 'SIL Specification',
render: () => `
<h1 class='page-title'>The Semantic Interoperability Layer (SIL)</h1>
<div class='card highlight'>
<div class='card-title'><span class='card-icon'>π§ </span>The Principal Theoretical Contribution</div>
<p style='margin:0;font-size:.95rem;color:var(--text)'>SIL is a normative, evidence-mapping function that takes heterogeneous disability evidence and produces a claimant-specific, adjudicator-legible representation of functional impairment and its work-limiting effects. It is <strong>not</strong> a database, a vocabulary, or a checklist. It is an <strong>interpretive function</strong> β closer in form to a latent-variable model than to a data schema.</p>
</div>
<h2 class='section-heading'>Formal Specification</h2>
<p>Notation:</p>
<div class='math-block'>
<span class='comment'>// Let i = 1,β¦,n index pieces of evidence for a given claimant</span><br>
<span class='comment'>// Let j = 1,β¦,m index SSA-relevant functional domains</span><br><br>
E<sub>i</sub> = evidence piece i (type, source, date, content, coded fields)<br>
X<sub>j</sub> = latent functional capacity in domain j (0β1 scale)<br>
X = (X<sub>1</sub>, β¦, X<sub>m</sub>)<sup>β€</sup> = vector of latent capacities
</div>
<h3 class='sub-heading'>Measurement Model</h3>
<div class='math-block'>
f<sub>i</sub>(E<sub>i</sub>) = Ξ<sub>i</sub><sup>β€</sup>X + Ξ΅<sub>i</sub><br><br>
<span class='comment'>// f_i extracts measurement-relevant signal from evidence E_i</span><br>
<span class='comment'>// Ξ_i = vector of loadings (zero for domains evidence doesn't speak to)</span><br>
<span class='comment'>// Ξ΅_i = evidence-specific residual</span><br><br>
<span class='comment'>// Two mapping routes:</span><br>
<span class='comment'>// 1. Rule-based β for evidence with established mappings</span><br>
<span class='comment'>// 2. Learned β for evidence without (genomic, imaging, novel measures)</span>
</div>
<h3 class='sub-heading'>Structural Model</h3>
<div class='math-block'>
X = BW + ΞΆ<br><br>
<span class='comment'>// W = overall latent work-disability construct</span><br>
<span class='comment'>// B = vector of structural loadings</span><br>
<span class='comment'>// ΞΆ = domain-specific residuals</span><br><br>
<span class='comment'>// Optional: additional latent constructs (mental-health, physical,</span><br>
<span class='comment'>// cognitive-executive factors) as simultaneous equations (Bollen 1989)</span>
</div>
<h3 class='sub-heading'>Estimation</h3>
<p style='font-size:.85rem'>Parameters are estimated by minimizing the discrepancy between model-implied and observed cross-evidence consistency matrices. The observed consistency matrix is constructed by pairing each evidence piece with each other piece, coding consistency (concordant / complementary / conflicting / unrelated), and aggregating. Estimation uses diagonally weighted least squares (DWLS) β the same estimator as Genomic-SEM β adapted to ordinal consistency coding.</p>
<h3 class='sub-heading'>Downstream Output</h3>
<div class='table-wrap'>
<table>
<thead><tr><th>Output</th><th>Description</th></tr></thead>
<tbody>
<tr><td><strong>Ε΄ score</strong></td><td>Estimated work-disability score with confidence interval</td></tr>
<tr><td><strong>XΜ vector</strong></td><td>Domain-by-domain capacity scores with CIs</td></tr>
<tr><td><strong>Evidence map</strong></td><td>Machine-readable trace from each claim to underlying source and factor loading</td></tr>
<tr><td><strong>Consistency report</strong></td><td>Flags concordant, complementary, and conflicting evidence</td></tr>
</tbody>
</table>
</div>
<h2 class='section-heading'>What SIL Is Not</h2>
<div style='display:grid;grid-template-columns:1fr 1fr;gap:12px'>
<div class='card'><p style='margin:0;font-size:.82rem'><strong>β</strong> A decision-making system β SIL does not allow or deny benefits. The adjudicator decides.</p></div>
<div class='card'><p style='margin:0;font-size:.82rem'><strong>β</strong> A black box β Every SIL output traces back to underlying evidence and loadings.</p></div>
<div class='card'><p style='margin:0;font-size:.82rem'><strong>β</strong> A replacement for expertise β SIL is a common representation of clinical and vocational inputs.</p></div>
<div class='card'><p style='margin:0;font-size:.82rem'><strong>β</strong> A one-time model β SIL is per-claimant instantiation of a model class; parameters update.</p></div>
</div>
`
};
sections.analogy = {
title: 'Genomic-SEM Analogy',
render: () => `
<h1 class='page-title'>SIL β Genomic-SEM: A Structural Parallel</h1>
<p class='lead'>The analogy is deliberately drawn β and we argue it is genuinely illuminating rather than ornamental.</p>
<div class='table-wrap'>
<table>
<thead><tr><th>Component</th><th>Genomic-SEM (Grotzinger 2019)</th><th>SIL (This Paper)</th></tr></thead>
<tbody>
<tr><td><strong>Federated evidence</strong></td><td>Multiple GWAS cohorts with different phenotypes, instruments, sample sizes</td><td>Medical encounters, CE reports, work history, genomic findings, imaging, lay testimony</td></tr>
<tr><td><strong>Covariance object</strong></td><td>Genetic covariance matrix (via LD score regression)</td><td>Cross-evidence consistency matrix (via paired consistency coding)</td></tr>
<tr><td><strong>Latent construct</strong></td><td>Common genetic factor explaining shared signal across traits</td><td>Functional-impairment factor explaining shared signal across evidence</td></tr>
<tr><td><strong>Downstream output</strong></td><td>Multivariate GWAS on latent factor</td><td>Adjudicator-readable evidence map</td></tr>
<tr><td><strong>Estimator</strong></td><td>DWLS on genetic covariance</td><td>DWLS on consistency matrix (adapted to ordinal coding)</td></tr>
</tbody>
</table>
</div>
<div class='card highlight'>
<div class='card-title'><span class='card-icon'>π</span>Why the Analogy Works</div>
<p>Both operate on federated evidence β data never collected in a common model. Both construct a covariance-like object from pairwise relationships. Both estimate a latent construct that no individual data point directly measures. Both produce a downstream representation usable by the relevant community (statistical geneticists β adjudicators).</p>
<p style='margin-bottom:0'>The analogies are not perfect β disability evidence is much messier than GWAS summary statistics β but the <strong>structural parallel is real</strong>, and importing the latent-variable formalism is more than cosmetic.</p>
</div>
`
};
sections.rubric = {
title: 'Scoring Rubric',
render: () => `
<h1 class='page-title'>Exhibit A β Four-Level Scoring Rubric</h1>
<p class='lead'>Each level scored 0β3 against five criteria β per-level 0β15 β total 0β60.</p>
<div class='rubric-calculator' id='rubricCalc'>
<div class='card-title' style='margin-bottom:12px'>ποΈ Interactive Rubric Calculator</div>
<div class='rubric-controls' id='rubricControls'>
<button class='rubric-btn active' data-module='psdlef'>PSDLEF (Overall)</button>
<button class='rubric-btn' data-module='aidoc'>AI Doc Gen</button>
<button class='rubric-btn' data-module='neuro'>Neurodiversity + Genomic</button>
<button class='rubric-btn' data-module='court'>Court Navigation</button>
</div>
<div class='table-wrap' style='margin:12px 0 0'>
<table id='rubricTable'>
<thead><tr><th>Criterion</th><th>Description</th><th>Score (0β3)</th></tr></thead>
<tbody id='rubricBody'></tbody>
</table>
</div>
<div class='rubric-score-display' id='rubricScoreDisplay'></div>
</div>
<h2 class='section-heading'>Interpretation Bands</h2>
<div class='table-wrap'>
<table>
<thead><tr><th>Score Range</th><th>Interpretation</th></tr></thead>
<tbody>
<tr><td class='score-cell score-0'>0β15</td><td>System essentially non-interoperable; cannot support meaningful adjudication</td></tr>
<tr><td class='score-cell score-1'>16β30</td><td>Foundational and partial structural only; semantic and organizational absent</td></tr>
<tr><td class='score-cell score-2'>31β45</td><td>Functional for straightforward cases; fails on complex, multimodal evidence</td></tr>
<tr><td class='score-cell score-3'>46β60</td><td>Capable of supporting full SIL-based pro se litigation model</td></tr>
</tbody>
</table>
</div>
`,
init: () => initRubric()
};
const rubricData = {
psdlef: {
name: 'PSDLEF (Overall)', scores: [
['L1.1 Connectivity', 'One-way upload; some auth', 1],
['L1.2 Identity', 'MFA implemented', 2],
['L1.3 Consent', 'Digital, revocable', 2],
['L1.4 Reliability', 'Designed SLA', 2],
['L1.5 Latency', 'Real-time for routine ops', 2],
['L2.1 Schema', 'Template-based, partial', 1],
['L2.2 Standards', 'No USCDI/FHIR adoption', 0],
['L2.3 Imaging/Genomic', 'Ad hoc wrappers', 1],
['L2.4 Backward compat', 'Documented migration', 1],
['L2.5 Inter-system map', 'Partner-documented', 1],
['L3.1 Vocabularies', 'Local codes only', 0],
['L3.2 Functional mapping', 'Implicit in narrative', 1],
['L3.3 Consistency', 'Manual review only', 0],
['L3.4 Provenance', 'Partial trail', 1],
['L3.5 Adjudicator legibility', 'Specialist only', 0],
['L4.1 Governance', 'Internal policies', 1],
['L4.2 Consent regime', 'Digital, granular', 2],
['L4.3 Adjudicator engagement', 'Not addressed', 0],
['L4.4 Remedy pathway', 'Internal grievance', 1],
['L4.5 Equity', 'Designed-in', 2]
]
},
aidoc: {
name: 'AI Doc Gen', scores: [
['L1.1 Connectivity', 'Web portal, no SSA fetch', 2],
['L1.2 Identity', 'MFA', 2],
['L1.3 Consent', 'Digital, revocable', 2],
['L1.4 Reliability', 'Designed SLA', 2],
['L1.5 Latency', 'Real-time doc gen', 2],
['L2.1 Schema', 'Template-based', 1],
['L2.2 Standards', 'None', 0],
['L2.3 Imaging/Genomic', 'None', 0],
['L2.4 Backward compat', 'Basic', 1],
['L2.5 Inter-system map', 'None', 0],
['L3.1 Vocabularies', 'Legal English only', 0],
['L3.2 Functional mapping', 'None', 0],
['L3.3 Consistency', 'None', 0],
['L3.4 Provenance', 'Per-template', 1],
['L3.5 Adjudicator legibility', 'Legal prose', 1],
['L4.1 Governance', 'Internal', 1],
['L4.2 Consent regime', 'Digital', 2],
['L4.3 Adjudicator engagement', 'Not addressed', 0],
['L4.4 Remedy pathway', 'Grievance', 1],
['L4.5 Equity', 'Designed', 2]
]
},
neuro: {
name: 'Neurodiversity + Genomic', scores: [
['L1.1 Connectivity', 'Intake portal, no EHR', 1],
['L1.2 Identity', 'MFA', 2],
['L1.3 Consent', 'Granular, auditable', 3],
['L1.4 Reliability', 'Not specified', 1],
['L1.5 Latency', 'Asynchronous', 1],
['L2.1 Schema', 'Free text + JSON', 1],
['L2.2 Standards', 'HL7 Genomics where applicable', 1],
['L2.3 Imaging/Genomic', 'VCF, DICOM wrappers', 2],
['L2.4 Backward compat', 'Basic', 1],
['L2.5 Inter-system map', 'EHR partners', 1],
['L3.1 Vocabularies', 'SNOMED, HPO', 2],
['L3.2 Functional mapping', 'Implicit', 1],
['L3.3 Consistency', 'Manual review', 1],
['L3.4 Provenance', 'Per-variant, per-claim', 2],
['L3.5 Adjudicator legibility', 'No SIL', 0],
['L4.1 Governance', 'Consent regime only', 1],
['L4.2 Consent regime', 'Granular', 3],
['L4.3 Adjudicator engagement', 'Not addressed', 0],
['L4.4 Remedy pathway', 'Grievance', 1],
['L4.5 Equity', 'Designed + neurodivergent', 3]
]
},
court: {
name: 'Court Navigation', scores: [
['L1.1 Connectivity', 'Web + SMS, PACER', 2],
['L1.2 Identity', 'Lightweight', 1],
['L1.3 Consent', 'Basic', 1],
['L1.4 Reliability', 'Designed', 2],
['L1.5 Latency', 'Real-time', 2],
['L2.1 Schema', 'Template', 1],
['L2.2 Standards', 'None', 0],
['L2.3 Imaging/Genomic', 'None', 0],
['L2.4 Backward compat', 'Basic', 1],
['L2.5 Inter-system map', 'PACER only', 1],
['L3.1 Vocabularies', 'None', 0],
['L3.2 Functional mapping', 'None', 0],
['L3.3 Consistency', 'None', 0],
['L3.4 Provenance', 'None', 0],
['L3.5 Adjudicator legibility', 'Procedural prose', 1],
['L4.1 Governance', 'Internal', 1],
['L4.2 Consent regime', 'Basic', 1],
['L4.3 Adjudicator engagement', 'Not addressed', 0],
['L4.4 Remedy pathway', 'Grievance', 1],
['L4.5 Equity', 'Designed', 2]
]
}
};
function initRubric(){
const controls = document.getElementById('rubricControls');
const body = document.getElementById('rubricBody');
const display = document.getElementById('rubricScoreDisplay');
let currentModule = 'psdlef';
function renderRubric(moduleKey){
const data = rubricData[moduleKey];
if(!data) return;
body.innerHTML = data.scores.map(([criterion, desc, score]) => `
<tr>
<td style='font-weight:600;color:var(--text);font-size:.78rem'>${criterion}</td>
<td style='font-size:.78rem'>${desc}</td>
<td class='score-cell score-${score}'>${score}</td>
</tr>
`).join('');
const perLevel = {1:0,2:0,3:0,4:0};
data.scores.forEach(([criterion,,score], idx) => {
perLevel[Math.floor(idx/5)+1] += score;
});
const total = Object.values(perLevel).reduce((a,b)=>a+b,0);
let band = '';
if(total <= 15){band = 'Non-interoperable';}
else if(total <= 30){band = 'Foundational/Partial Structural';}
else if(total <= 45){band = 'Functional for Straightforward Cases';}
else {band = 'SIL-Capable';}
const bandColor = total <= 15 ? '#f85149' : total <= 30 ? '#f0883e' : total <= 45 ? '#58a6ff' : '#7ee787';
display.innerHTML = `
<div>
<span style='font-size:.72rem;color:var(--text3);display:block;margin-bottom:2px'>TOTAL SCORE</span>
<span class='rubric-total'>${total}<span style='font-size:1rem;color:var(--text3)'>/60</span></span>
</div>
<div style='flex:1;min-width:200px'>
<span style='font-size:.72rem;color:var(--text3);display:block;margin-bottom:4px'>PER-LEVEL BREAKDOWN</span>
<div style='display:flex;gap:4px;height:28px;align-items:flex-end'>
${[1,2,3,4].map(l => {
const pct = (perLevel[l]/15)*100;
const colors = {1:'#58a6ff',2:'#7ee787',3:'#f0883e',4:'#f85149'};
return `<div style='flex:1;display:flex;flex-direction:column;align-items:center;gap:2px'>
<span style='font-size:.62rem;color:var(--text2);font-weight:600'>${perLevel[l]}</span>
<div style='width:100%;background:var(--bg3);border-radius:4px;height:10px'><div style='width:${pct}%;background:${colors[l]};height:100%;border-radius:4px;transition:width .8s ease'></div></div>
<span style='font-size:.56rem;color:var(--text3)'>L${l}</span>
</div>`;
}).join('')}
</div>
</div>
<div>
<span style='font-size:.72rem;color:var(--text3);display:block;margin-bottom:2px'>BAND</span>
<span class='rubric-band' style='background:${bandColor}22;color:${bandColor};border:1px solid ${bandColor}44'>${band}</span>
</div>
`;
}
controls.querySelectorAll('.rubric-btn').forEach(btn => {
btn.addEventListener('click', () => {
controls.querySelectorAll('.rubric-btn').forEach(b => b.classList.remove('active'));
btn.classList.add('active');
currentModule = btn.dataset.module;
renderRubric(currentModule);
});
});
renderRubric(currentModule);
}
sections['worked-example'] = {
title: 'Worked Example',
render: () => `
<h1 class='page-title'>Exhibit B β Worked Example</h1>
<p class='lead'>PSDLEF litigant-services modules scored against the four-level rubric.</p>
<div class='chart-container'>
<div class='chart-title'>Module Scores by Level (0β15 per level)</div>
<div style='display:grid;grid-template-columns:repeat(3,1fr);gap:16px' id='moduleCharts'></div>
</div>
<div class='table-wrap'>
<table>
<thead><tr><th>Module</th><th>L1 Foundational</th><th>L2 Structural</th><th>L3 Semantic</th><th>L4 Organizational</th><th>Total /60</th><th>Interpretation</th></tr></thead>
<tbody>
<tr>
<td><strong>AI Doc Gen</strong></td>
<td class='score-cell score-2'>10</td>
<td class='score-cell score-0'>2</td>
<td class='score-cell score-0'>2</td>
<td class='score-cell score-1'>6</td>
<td class='score-cell'>20</td>
<td style='font-size:.74rem;color:var(--warn)'>Below functional band</td>
</tr>
<tr>
<td><strong>Neurodiversity + Genomic</strong></td>
<td class='score-cell score-2'>8</td>
<td class='score-cell score-1'>6</td>
<td class='score-cell score-1'>6</td>
<td class='score-cell score-2'>8</td>
<td class='score-cell'>28</td>
<td style='font-size:.74rem;color:var(--warn)'>Below functional band β strongest overall</td>
</tr>
<tr>
<td><strong>Court Navigation</strong></td>
<td class='score-cell score-2'>8</td>
<td class='score-cell score-0'>3</td>
<td class='score-cell score-0'>1</td>
<td class='score-cell score-1'>5</td>
<td class='score-cell'>17</td>
<td style='font-size:.74rem;color:var(--warn)'>Below functional band β weakest</td>
</tr>
</tbody>
</table>
</div>
<div class='card highlight'>
<div class='card-title'><span class='card-icon'>π―</span>Highest-Leverage Investments for PSDLEF</div>
<ol style='padding-left:20px;color:var(--text2);font-size:.85rem;line-height:1.8'>
<li><strong>Build SIL</strong> β Integrate the latent-factor functional-impairment model into the neurodiversity + genomic module first (where the semantic gap is widest).</li>
<li><strong>Negotiate the adjudicator-engagement protocol</strong> (Level 4) β Without this, the semantic layer has no institutional home.</li>
<li><strong>Define and publish the PSDLEF Evidence Schema</strong> (Level 2) β Schema publication creates a reference for partners and benchmarks future adoption.</li>
</ol>
</div>
`,
init: () => initModuleCharts()
};
function initModuleCharts(){
const container = document.getElementById('moduleCharts');
if(!container) return;
const data = [
{name:'AI Doc Gen', levels:[10,2,2,6], total:20, color:'#58a6ff'},
{name:'Neuro + Genomic', levels:[8,6,6,8], total:28, color:'#7ee787'},
{name:'Court Navigation', levels:[8,3,1,5], total:17, color:'#f0883e'}
];
container.innerHTML = data.map((mod,i) => {
const maxPerLevel = 15;
return `
<div style='background:var(--bg3);border-radius:10px;padding:16px;text-align:center'>
<div style='font-size:.74rem;font-weight:700;margin-bottom:10px;color:var(--text)'>${mod.name}</div>
<div style='font-size:1.3rem;font-weight:800;font-family:var(--mono);color:${mod.color};margin-bottom:10px'>${mod.total}/60</div>
${[1,2,3,4].map(l => {
const pct = (mod.levels[l-1]/maxPerLevel)*100;
return `<div style='display:flex;align-items:center;gap:6px;margin-bottom:4px'>
<span style='font-size:.56rem;color:var(--text3);width:16px'>L${l}</span>
<div style='flex:1;background:var(--bg);border-radius:4px;height:8px'><div style='width:${pct}%;background:${mod.color};height:100%;border-radius:4px;transition:width 1s ease ${i*200 + l*100}ms'></div></div>
<span style='font-size:.56rem;color:var(--text2);width:16px'>${mod.levels[l-1]}</span>
</div>`;
}).join('')}
</div>
`;
}).join('');
requestAnimationFrame(() => {
setTimeout(() => {
container.querySelectorAll('[style*="width:0%"]').forEach(el => {
const style = el.getAttribute('style');
const match = style.match(/width:(\d+)%/);
if(match){
el.style.width = match[1] + '%';
el.style.transition = 'width .8s ease';
}
});
}, 150);
});
}
sections.implications = {
title: 'Implications',
render: () => `
<h1 class='page-title'>Implications</h1>
<h2 class='section-heading'>For PSDLEF Program Design</h2>
<div class='card'>
<p style='margin:0'><strong>Build SIL before scaling the AI doc gen module.</strong> Doc gen without SIL produces documents that are pro se-formatted but not semantically interoperable β they reproduce the lay vocabulary of the claimant rather than translating into the adjudicator's functional-impairment frame.</p>
</div>
<div class='card'>
<p style='margin:0'><strong>Use the rubric as a portfolio instrument.</strong> The PSDLEF Board can evaluate proposed modules, partner integrations, and grant applications. A partner that does not raise at least Level 1 and Level 4 scores should not be assumed to add semantic value.</p>
</div>
<div class='card'>
<p style='margin:0'><strong>Engage SSA on Levels 2 and 4 before Level 3.</strong> SIL is technically feasible today; deployment is gated by organizational interoperability, not technical capability. The political-organisational engagement is the binding constraint.</p>
</div>
<h2 class='section-heading'>For Pro Se Litigation Practice</h2>
<div class='callout accent'>
<span class='callout-icon'>π©Ί</span>
<div class='callout-content'><p>The four-level model provides a <strong>diagnostic vocabulary</strong> for case-level failures: delay = Level 1 (foundational); records misinterpreted = Level 3 (semantic); valid claim denied three times = Level 4 (organizational). A pro se litigant can identify <em>which level is failing</em> and target the remedy accordingly β a forensic use of the framework.</p></div>
</div>
<h2 class='section-heading'>Limitations</h2>
<ul style='padding-left:20px;color:var(--text2);font-size:.85rem;line-height:1.8'>
<li>Theoretical β borrowed from health IT, not empirically validated in disability adjudication.</li>
<li>SIL formal model at abstraction level comparable to Genomic-SEM spec; estimators and validation are future work.</li>
<li>Rubric proposed, not validated; worked example illustrative.</li>
<li>Does not address non-interoperability drivers: understaffing, ALJ hiring, judicial economy, political economy of benefits.</li>
</ul>
<h2 class='section-heading'>Future Research</h2>
<ol style='padding-left:20px;color:var(--text2);font-size:.85rem;line-height:1.8'>
<li>Empirical validation of the rubric against pro se disability cases with inter-rater reliability.</li>
<li>SIL pilot on retrospective adjudicated claims, comparing SIL-derived scores to actual outcomes.</li>
<li>Adjudicator-engagement protocol design with SSA, OHO, and federal courts.</li>
<li>Standards engagement with HL7, ONC, HIMSS on disability-specific interoperability profiles.</li>
<li>Cross-jurisdictional comparison: UK PIP, German DRV, Australian DSP.</li>
</ol>
`
};
sections.references = {
title: 'References',
render: () => `
<h1 class='page-title'>References</h1>
<div style='display:flex;flex-direction:column;gap:10px'>
${[
['Bollen, K. A. (1989). <em>Structural Equations with Latent Variables</em>. Wiley.', ''],
['Gelber, A., Moore, T., Pei, Z., & Strand, A. (2023). Disability insurance income saves lives. <em>Journal of Political Economy</em>, 131(11), 3023β3075.', ''],
['Grotzinger, A. D., Rhemtulla, M., de Vlaming, R., et al. (2019). Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. <em>Nature Human Behaviour</em>, 3(5), 513β525.', 'tag purple'],
['HIMSS. (2024). <em>Interoperability in Healthcare</em>. Healthcare Information and Management Systems Society.', 'tag accent'],
['HIMSS. (2024). <em>HIMSS Interoperability Frameworks eBook</em>.', 'tag accent'],
['Kenya Ministry of Health. (2019). <em>Kenya Health Information Systems Interoperability Framework (KHISIF)</em>.', ''],
['McLaughlin, C. L. (2026). <em>Research Portfolio Exhibit: SSA Deaths While Pending, Administrative Recidivism, and the Prison-Industrial Feedback Loop</em>. (Litigation-support exhibit, prepared pro se.)', 'tag orange'],
['ONC. (2024). <em>TEFCA: Trusted Exchange Framework and Common Agreement</em>. Office of the National Coordinator for Health IT.', 'tag accent'],
['PSDLEF. (2026). <em>PSDLEF: A Perpetual Framework for Pro Se Disability Litigant Empowerment Through Algorithmic Capital Stewardship and Sustainable Municipal Investment</em>. Working paper.', 'tag green'],
['U.S. GAO. (2020). <em>Social Security Disability: Information on Wait Times, Bankruptcies, and Deaths among Applicants Who Appealed Benefit Denials</em> (GAO-20-641R).', ''],
['Walker, J., Pan, E., Johnston, D., et al. (2005). The value of health care information exchange and interoperability. <em>Health Affairs</em>, 24(S1), W5-10βW5-18.', ''],
['Wolters Kluwer. (2024). <em>Understanding the Four Levels of Interoperability in Healthcare</em>.', 'tag accent']
].map(([ref, tag]) => `
<div class='card' style='padding:14px 18px;margin-bottom:0'>
${tag ? `<span class='tag ${tag}' style='font-size:.6rem'>${tag.includes('purple')?'Genomic-SEM':tag.includes('accent')?'HIMSS/ONC':tag.includes('orange')?'Exhibit':tag.includes('green')?'PSDLEF':''}</span>` : ''}
<p style='margin:0;font-size:.82rem'>${ref}</p>
</div>
`).join('')}
</div>
<h2 class='section-heading'>Appendices</h2>
<div class='card'>
<div class='card-title'>Appendix A β HIMSS Level Definitions (Verbatim)</div>
<p style='font-size:.78rem;color:var(--text2);margin-bottom:6px'><strong>Level 1 β Foundational.</strong> Establishes inter-connectivity requirements for systems to securely communicate data to and receive data from another.</p>
<p style='font-size:.78rem;color:var(--text2);margin-bottom:6px'><strong>Level 2 β Structural.</strong> Defines format, syntax, and organization of data exchange (including at data-field level) for interpretation by the receiving system.</p>
<p style='font-size:.78rem;color:var(--text2);margin-bottom:6px'><strong>Level 3 β Semantic.</strong> Provides for common underlying models and codification of data, including data elements with standardized definitions from publicly available value sets, providing shared understanding and meaning.</p>
<p style='font-size:.78rem;color:var(--text2)'><strong>Level 4 β Organizational.</strong> Includes governance, policy, social, legal, and organizational considerations to facilitate secure, seamless, timely communication and use of data. Enables shared consent, trust, and integrated workflows.</p>
</div>
`
};
// ββ Router / AJAX Simulation βββββββββββββββββββββββββββββββββββββ
const sectionNames = {
'abstract': 'Abstract',
'introduction': 'Introduction',
'four-levels': 'Four Levels',
'diagnostic': 'Diagnostic Analysis',
'sil': 'SIL Specification',
'analogy': 'Genomic-SEM Analogy',
'rubric': 'Scoring Rubric',
'worked-example': 'Worked Example',
'implications': 'Implications',
'references': 'References'
};
let currentSection = 'abstract';
function navigateTo(sectionKey, updateHash = true){
if(!sections[sectionKey]) return;
const container = document.getElementById('sectionContainer');
const skeleton = document.getElementById('loadingSkeleton');
currentSection = sectionKey;
// Show loading skeleton (simulating AJAX)
container.style.display = 'none';
skeleton.style.display = 'block';
// Simulate network delay for AJAX feel
setTimeout(() => {
skeleton.style.display = 'none';
container.style.display = 'block';
container.innerHTML = sections[sectionKey].render();
if(sections[sectionKey].init) sections[sectionKey].init();
// Update breadcrumb
document.getElementById('breadcrumbCurrent').textContent = sectionNames[sectionKey] || sectionKey;
// Update nav
document.querySelectorAll('.nav-link').forEach(link => {
link.classList.toggle('active', link.dataset.section === sectionKey);
});
// Scroll to top of content
document.querySelector('.content').scrollIntoView({behavior: 'smooth', block: 'start'});
if(updateHash) {
history.pushState(null, '', '#' + sectionKey);
}
}, 250 + Math.random() * 200);
}
// ββ Event Listeners ββββββββββββββββββββββββββββββββββββββββββββββ
document.querySelectorAll('.nav-link').forEach(link => {
link.addEventListener('click', () => {
navigateTo(link.dataset.section);
// Close mobile sidebar
document.getElementById('sidebar').classList.remove('open');
});
});
// Hash routing
window.addEventListener('hashchange', () => {
const hash = window.location.hash.replace('#', '');
if(hash && sections[hash] && hash !== currentSection) {
navigateTo(hash, false);
}
});
// Hamburger
document.getElementById('hamburger').addEventListener('click', () => {
document.getElementById('sidebar').classList.toggle('open');
});
// Theme toggle (placeholder for light theme)
document.getElementById('themeToggle').addEventListener('click', () => {
const root = document.documentElement;
if(root.style.getPropertyValue('--bg') === '#f5f5f5'){
root.style.setProperty('--bg','#0d1117');
root.style.setProperty('--bg2','#161b22');
root.style.setProperty('--bg3','#1c2430');
root.style.setProperty('--card','#1a2029');
root.style.setProperty('--card2','#202834');
root.style.setProperty('--text','#e6edf3');
root.style.setProperty('--text2','#9da7b3');
root.style.setProperty('--text3','#6e7a8a');
root.style.setProperty('--border','#30363d');
root.style.setProperty('--border2','#3d4753');
} else {
root.style.setProperty('--bg','#f5f5f5');
root.style.setProperty('--bg2','#e8e8e8');
root.style.setProperty('--bg3','#ddd');
root.style.setProperty('--card','#fff');
root.style.setProperty('--card2','#f0f0f0');
root.style.setProperty('--text','#1a1a1a');
root.style.setProperty('--text2','#555');
root.style.setProperty('--text3','#888');
root.style.setProperty('--border','#ccc');
root.style.setProperty('--border2','#bbb');
}
});
// Search
document.getElementById('searchInput').addEventListener('input', (e) => {
const query = e.target.value.toLowerCase().trim();
const results = Object.keys(sectionNames).filter(key => {
const name = sectionNames[key].toLowerCase();
const content = sections[key]?.render?.()?.toLowerCase() || '';
return name.includes(query) || content.includes(query);
});
// Highlight matching nav links
document.querySelectorAll('.nav-link').forEach(link => {
const key = link.dataset.section;
if(query && results.includes(key)){
link.style.outline = '1px solid var(--accent)';
link.style.outlineOffset = '-1px';
} else {
link.style.outline = 'none';
}
});
// If exact match, navigate
if(query && results.length === 1 && query.length > 3){
// Don't auto-navigate, just show
}
});
// Modal close
document.getElementById('modalClose').addEventListener('click', () => {
document.getElementById('modalOverlay').classList.remove('visible');
});
document.getElementById('modalOverlay').addEventListener('click', (e) => {
if(e.target === e.target.closest('.modal-overlay')){
document.getElementById('modalOverlay').classList.remove('visible');
}
});
// ββ Initial Load βββββββββββββββββββββββββββββββββββββββββββββββββ
const initialHash = window.location.hash.replace('#','');
navigateTo(initialHash && sections[initialHash] ? initialHash : 'abstract', false);
// Keyboard shortcuts
document.addEventListener('keydown', (e) => {
if(e.target.tagName === 'INPUT') return;
const keys = Object.keys(sectionNames);
const idx = keys.indexOf(currentSection);
if(e.key === 'ArrowRight' && idx < keys.length - 1){
navigateTo(keys[idx + 1]);
}
if(e.key === 'ArrowLeft' && idx > 0){
navigateTo(keys[idx - 1]);
}
if(e.key === 'Escape'){
document.getElementById('modalOverlay').classList.remove('visible');
document.getElementById('sidebar').classList.remove('open');
}
});
})();
</script>
</body>
</html>" style="width:100%;height:800px;border:2px solid #30363d;border-radius:12px;background:#0d1117">]]>