Datasets:

lukeslp
/

accessibility-atlas

+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Accessibility Atlas: A Data-Driven Portrait of Disability\n",
+    "\n",
+    "**Author**: Luke Steuber  \n",
+    "**Date**: February 2026  \n",
+    "**Data Sources**: US Census Bureau, Bureau of Labor Statistics, WebAIM, Eurostat, NCES/IDEA, and more  \n",
+    "\n",
+    "This notebook explores disability prevalence, employment outcomes, web accessibility compliance, assistive technology usage, and special education trends across 25+ datasets.\n",
+    "\n",
+    "---"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "import json\nimport csv\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport matplotlib.ticker as mticker\nfrom pathlib import Path\nimport warnings\nwarnings.filterwarnings('ignore')\n\n# Style setup\nplt.rcParams['figure.figsize'] = (12, 6)\nplt.rcParams['figure.dpi'] = 100\nplt.rcParams['axes.titlesize'] = 14\nplt.rcParams['axes.labelsize'] = 12\nplt.rcParams['axes.grid'] = True\nplt.rcParams['grid.alpha'] = 0.3\nplt.rcParams['font.family'] = 'sans-serif'\n\nDATA_DIR = Path('.')\njson_files = list(DATA_DIR.glob('*.json'))\ncsv_files = list(DATA_DIR.glob('*.csv'))\nprint(f'Data directory: {DATA_DIR.resolve()}')\nprint(f'Files: {len(json_files)} JSON + {len(csv_files)} CSV = {len(json_files) + len(csv_files)} datasets')\n\ndef read_csv_as_dicts(path):\n    \"\"\"Read CSV into list of dicts (no pandas needed).\"\"\"\n    with open(path, newline='', encoding='utf-8') as f:\n        return list(csv.DictReader(f))"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 1. US Disability Prevalence — National Trends (2010-2023)\n",
+    "\n",
+    "Census Bureau ACS 1-year estimates from Table S1810. 13 years of data (2020 excluded due to COVID survey disruptions)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "# Load the full 13-year Census trend dataset\nwith open(DATA_DIR / 'census_disability_trends_2010_2023.json') as f:\n    census_raw = json.load(f)\n\n# Parse S1810 into numpy arrays\n# Note: S1810 age-group variables changed meaning around 2015.\n# Pre-2015 values (0.4-0.8%) are NOT disability rates — skip those.\nt_years, t_pop, t_dis, t_pct = [], [], [], []\nt_u18, t_1864, t_65 = [], [], []\n\nfor year in sorted(census_raw['s1810_data'].keys()):\n    d = census_raw['s1810_data'][year]['data']\n    header, values = list(d[0]), list(d[1])\n    row = dict(zip(header, values))\n    yr = int(year)\n    \n    t_years.append(yr)\n    t_pop.append(float(row.get('S1810_C01_001E', 0) or 0))\n    t_dis.append(float(row.get('S1810_C02_001E', 0) or 0))\n    t_pct.append(float(row.get('S1810_C03_001E', 0) or 0))\n    \n    # Pre-2015 age-group data uses different variable definitions\n    if yr < 2015:\n        t_u18.append(np.nan)\n        t_1864.append(np.nan)\n        t_65.append(np.nan)\n    else:\n        t_u18.append(float(row.get('S1810_C03_002E', 0) or 0))\n        t_1864.append(float(row.get('S1810_C03_003E', 0) or 0))\n        t_65.append(float(row.get('S1810_C03_004E', 0) or 0))\n\n# Convert to numpy\nt_years = np.array(t_years)\nt_pop = np.array(t_pop)\nt_dis = np.array(t_dis)\nt_pct = np.array(t_pct)\nt_u18 = np.array(t_u18)\nt_1864 = np.array(t_1864)\nt_65 = np.array(t_65)\n\nprint(f'Census disability trends: {len(t_years)} years ({t_years[0]}-{t_years[-1]})')\nprint(f'Disability rate: {t_pct[0]:.1f}% ({t_years[0]}) → {t_pct[-1]:.1f}% ({t_years[-1]})')\nprint(f'Population with disability: {t_dis[-1]:,.0f} in {t_years[-1]}')"
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 6))\n\n# Left: Overall disability rate\nax1.plot(t_years, t_pct, 'o-', color='#2c3e50', linewidth=2.5, markersize=8)\nax1.fill_between(t_years, t_pct, alpha=0.1, color='#2c3e50')\nax1.set_title('US Disability Prevalence Rate', fontweight='bold')\nax1.set_ylabel('% of Population')\nax1.set_xlabel('Year')\nax1.axvspan(2019.5, 2020.5, alpha=0.15, color='red', label='2020 gap (COVID)')\nax1.legend()\nax1.yaxis.set_major_formatter(mticker.FormatStrFormatter('%.1f%%'))\n\n# Right: By age group (2015+ only, where data is valid)\nmask = ~np.isnan(t_u18)\nax2.plot(t_years[mask], t_u18[mask], 'o-', color='#3498db', label='Under 18', linewidth=2, markersize=6)\nax2.plot(t_years[mask], t_1864[mask], 'o-', color='#e67e22', label='18-64', linewidth=2, markersize=6)\nax2.plot(t_years[mask], t_65[mask], 'o-', color='#e74c3c', label='65+', linewidth=2, markersize=6)\n\nax2.set_title('Disability Rate by Age Group', fontweight='bold')\nax2.set_ylabel('% of Age Group')\nax2.set_xlabel('Year')\nax2.legend()\nax2.yaxis.set_major_formatter(mticker.FormatStrFormatter('%.1f%%'))\n\nplt.tight_layout()\nplt.show()\n\nprint(f'\\nKey finding: Disability rate rose from {t_pct[0]:.1f}% ({t_years[0]}) to {t_pct[-1]:.1f}% ({t_years[-1]})')\nprint(f'Total with disability in {t_years[-1]}: {t_dis[-1]:,.0f}')"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 2. Disability by Age and Sex (2022 Snapshot)\n",
+    "\n",
+    "Detailed breakdown from Census Table B18101 showing how disability rates vary dramatically by age and sex."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'census_disability_by_age_sex_2022.json') as f:\n",
+    "    age_sex = json.load(f)\n",
+    "\n",
+    "# Build comparison DataFrame\n",
+    "age_groups = ['under_5', '5_to_17', '18_to_34', '35_to_64', '65_to_74', '75_plus']\n",
+    "age_labels = ['Under 5', '5-17', '18-34', '35-64', '65-74', '75+']\n",
+    "\n",
+    "male_rates = [age_sex['male']['by_age'][ag]['rate_pct'] for ag in age_groups]\n",
+    "female_rates = [age_sex['female']['by_age'][ag]['rate_pct'] for ag in age_groups]\n",
+    "\n",
+    "x = np.arange(len(age_labels))\n",
+    "width = 0.35\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(12, 6))\n",
+    "bars1 = ax.bar(x - width/2, male_rates, width, label='Male', color='#3498db', alpha=0.85)\n",
+    "bars2 = ax.bar(x + width/2, female_rates, width, label='Female', color='#e74c3c', alpha=0.85)\n",
+    "\n",
+    "ax.set_ylabel('Disability Rate (%)')\n",
+    "ax.set_title('Disability Rate by Age Group and Sex (2022)', fontweight='bold')\n",
+    "ax.set_xticks(x)\n",
+    "ax.set_xticklabels(age_labels)\n",
+    "ax.legend()\n",
+    "ax.bar_label(bars1, fmt='%.1f%%', padding=3, fontsize=9)\n",
+    "ax.bar_label(bars2, fmt='%.1f%%', padding=3, fontsize=9)\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'Overall: Male {age_sex[\"male\"][\"rate_pct\"]:.1f}% vs Female {age_sex[\"female\"][\"rate_pct\"]:.1f}%')\n",
+    "print(f'Steepest climb: 35-64 → 65-74 (male: {male_rates[3]:.1f}% → {male_rates[4]:.1f}%, female: {female_rates[3]:.1f}% → {female_rates[4]:.1f}%)')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 3. Disability Employment Gap\n",
+    "\n",
+    "Bureau of Labor Statistics data showing employment outcomes for people with and without disabilities."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'bls_disability_employment_2024.json') as f:\n",
+    "    bls = json.load(f)\n",
+    "\n",
+    "# Extract historical employment-population ratio\n",
+    "emp_data = bls['historical_trends']['employment_population_ratio']['data']\n",
+    "years_emp = sorted(emp_data.keys())\n",
+    "with_dis = [emp_data[y]['with_disability'] for y in years_emp]\n",
+    "without_dis = [emp_data[y].get('without_disability') for y in years_emp]\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 6))\n",
+    "\n",
+    "# Left: Employment-population ratio trend\n",
+    "ax1.plot([int(y) for y in years_emp], with_dis, 'o-', color='#e74c3c', linewidth=2, label='With disability')\n",
+    "# Plot without_disability where available\n",
+    "valid_wo = [(int(y), v) for y, v in zip(years_emp, without_dis) if v is not None]\n",
+    "if valid_wo:\n",
+    "    ax1.plot([x[0] for x in valid_wo], [x[1] for x in valid_wo], 's--', color='#2ecc71', linewidth=2, label='Without disability')\n",
+    "ax1.set_title('Employment-Population Ratio (2009-2024)', fontweight='bold')\n",
+    "ax1.set_ylabel('% Employed')\n",
+    "ax1.set_xlabel('Year')\n",
+    "ax1.legend()\n",
+    "ax1.axvspan(2019.5, 2020.5, alpha=0.1, color='gray')\n",
+    "\n",
+    "# Right: 2024 comparison dashboard\n",
+    "stats = bls['overall_statistics']\n",
+    "metrics = ['employment_population_ratio', 'unemployment_rate', 'part_time_workers_percent', 'self_employed_percent']\n",
+    "metric_labels = ['Employment Ratio', 'Unemployment Rate', 'Part-Time Workers', 'Self-Employed']\n",
+    "dis_vals = [stats['with_disability'][m] for m in metrics]\n",
+    "nodis_vals = [stats['without_disability'][m] if stats['without_disability'][m] is not None else 0 for m in metrics]\n",
+    "\n",
+    "x = np.arange(len(metric_labels))\n",
+    "ax2.barh(x - 0.2, dis_vals, 0.35, label='With Disability', color='#e74c3c', alpha=0.85)\n",
+    "ax2.barh(x + 0.2, nodis_vals, 0.35, label='Without Disability', color='#2ecc71', alpha=0.85)\n",
+    "ax2.set_yticks(x)\n",
+    "ax2.set_yticklabels(metric_labels)\n",
+    "ax2.set_xlabel('Percentage (%)')\n",
+    "ax2.set_title('Employment Metrics (2024)', fontweight='bold')\n",
+    "ax2.legend()\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'Employment gap: {stats[\"without_disability\"][\"employment_population_ratio\"] - stats[\"with_disability\"][\"employment_population_ratio\"]:.1f} percentage points')\n",
+    "print(f'People with disabilities employed: {bls[\"population_overview\"][\"total_employed_with_disability_thousands\"]}K')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Employment by race/ethnicity\n",
+    "race_data = bls['by_race_ethnicity']\n",
+    "races = list(race_data['disability_prevalence'].keys())\n",
+    "races = [r for r in races if r != 'note']\n",
+    "\n",
+    "race_labels = [r.replace('_', ' ').title() for r in races]\n",
+    "prevalence = [race_data['disability_prevalence'][r] for r in races]\n",
+    "unemp_dis = [race_data['unemployment_rate_with_disability'][r] for r in races]\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 5))\n",
+    "\n",
+    "bars = ax1.barh(race_labels, prevalence, color=['#3498db', '#e74c3c', '#2ecc71', '#f39c12'])\n",
+    "ax1.set_xlabel('Prevalence (%)')\n",
+    "ax1.set_title('Disability Prevalence by Race/Ethnicity', fontweight='bold')\n",
+    "ax1.bar_label(bars, fmt='%.1f%%', padding=5)\n",
+    "\n",
+    "bars2 = ax2.barh(race_labels, unemp_dis, color=['#3498db', '#e74c3c', '#2ecc71', '#f39c12'])\n",
+    "ax2.set_xlabel('Unemployment Rate (%)')\n",
+    "ax2.set_title('Unemployment Rate (With Disability)', fontweight='bold')\n",
+    "ax2.bar_label(bars2, fmt='%.1f%%', padding=5)\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "source": "with open(DATA_DIR / 'fred_disability_employment.json') as f:\n    fred = json.load(f)\n\nfred_years = sorted(fred['annual_data'].keys())\nfred_yr_int = [int(y) for y in fred_years]\nfred_emp_dis = [fred['annual_data'][y].get('disability_employment_ratio') for y in fred_years]\nfred_emp_all = [fred['annual_data'][y].get('total_employment_ratio') for y in fred_years]\nfred_gap = [fred['annual_data'][y].get('employment_gap_pp') for y in fred_years]\n\nfig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 6))\n\n# Left: Employment-population ratio comparison\nax1.plot(fred_yr_int, fred_emp_all, 's-', color='#2ecc71', linewidth=2.5, markersize=7, label='Total civilian')\nvalid_dis = [(y, v) for y, v in zip(fred_yr_int, fred_emp_dis) if v is not None]\nax1.plot([x[0] for x in valid_dis], [x[1] for x in valid_dis], 'o-', color='#e74c3c', linewidth=2.5, markersize=7, label='With disability')\nax1.fill_between([x[0] for x in valid_dis], [x[1] for x in valid_dis], \n                 [fred_emp_all[fred_yr_int.index(x[0])] for x in valid_dis], alpha=0.15, color='#e74c3c')\nax1.set_title('Employment-Population Ratio (FRED)', fontweight='bold')\nax1.set_ylabel('% Employed')\nax1.set_xlabel('Year')\nax1.legend()\nax1.axvspan(2019.5, 2020.5, alpha=0.1, color='gray', label='COVID')\n\n# Right: Gap over time\nvalid_gap = [(y, g) for y, g in zip(fred_yr_int, fred_gap) if g is not None]\nax2.bar([x[0] for x in valid_gap], [x[1] for x in valid_gap], color='#e67e22', alpha=0.7)\nax2.set_title('Employment Gap (pp)', fontweight='bold')\nax2.set_ylabel('Percentage Point Gap')\nax2.set_xlabel('Year')\nax2.axhline(y=np.mean([x[1] for x in valid_gap]), color='gray', linestyle='--', label=f'Average: {np.mean([x[1] for x in valid_gap]):.1f}pp')\nax2.legend()\n\nplt.tight_layout()\nplt.show()\n\ns = fred['summary']\nprint(f'FRED data: {s[\"years_of_data\"]} years (2009-{s[\"latest_year\"]})')\nprint(f'Employment gap ({s[\"latest_year\"]}): {s[\"employment_gap\"]} pp ({s[\"total_employment_ratio\"]}% total vs {s[\"disability_employment_ratio\"]}% disability)')\nprint(f'Disability LFPR: {s[\"disability_lfpr\"]}%')\nprint(f'{s[\"trend\"]}')",
+   "metadata": {},
+   "execution_count": null,
+   "outputs": []
+  },
+  {
+   "cell_type": "markdown",
+   "source": "### FRED: Disability Employment Gap Over Time\n\nComplete FRED series (2008-2024) showing the employment-population ratio for people with and without disabilities side by side.",
+   "metadata": {}
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 4. Web Accessibility — WebAIM Million Report\n",
+    "\n",
+    "WAVE automated analysis of 1,000,000 website home pages. The annual state-of-the-web for accessibility."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'webaim_million_2025.json') as f:\n",
+    "    webaim = json.load(f)\n",
+    "\n",
+    "trends = webaim['yearly_trends']\n",
+    "years = trends['years']\n",
+    "\n",
+    "fig, axes = plt.subplots(2, 2, figsize=(16, 12))\n",
+    "\n",
+    "# Top-left: Failure rate trend\n",
+    "ax = axes[0, 0]\n",
+    "ax.plot(years, trends['pages_with_failures_pct'], 'o-', color='#e74c3c', linewidth=2.5)\n",
+    "ax.fill_between(years, trends['pages_with_failures_pct'], 90, alpha=0.1, color='#e74c3c')\n",
+    "ax.set_title('Pages with WCAG Failures (%)', fontweight='bold')\n",
+    "ax.set_ylim(90, 100)\n",
+    "ax.yaxis.set_major_formatter(mticker.FormatStrFormatter('%.1f%%'))\n",
+    "ax.annotate(f'{trends[\"pages_with_failures_pct\"][-1]}%', xy=(years[-1], trends['pages_with_failures_pct'][-1]),\n",
+    "            fontsize=14, fontweight='bold', color='#e74c3c', ha='center', va='bottom',\n",
+    "            xytext=(0, 10), textcoords='offset points')\n",
+    "\n",
+    "# Top-right: Error types breakdown\n",
+    "ax = axes[0, 1]\n",
+    "error_types = ['low_contrast_pct', 'missing_alt_text_pct', 'empty_links_pct', \n",
+    "               'missing_form_labels_pct', 'empty_buttons_pct', 'missing_language_pct']\n",
+    "error_labels = ['Low Contrast', 'Missing Alt Text', 'Empty Links', \n",
+    "                'Missing Labels', 'Empty Buttons', 'Missing Lang']\n",
+    "colors_err = ['#e74c3c', '#3498db', '#f39c12', '#9b59b6', '#2ecc71', '#1abc9c']\n",
+    "for et, label, color in zip(error_types, error_labels, colors_err):\n",
+    "    ax.plot(years, trends[et], 'o-', label=label, color=color, linewidth=1.5)\n",
+    "ax.set_title('WCAG Failure Types Over Time', fontweight='bold')\n",
+    "ax.set_ylabel('% of Pages Affected')\n",
+    "ax.legend(fontsize=9, loc='upper right')\n",
+    "\n",
+    "# Bottom-left: CMS performance\n",
+    "ax = axes[1, 0]\n",
+    "cms_data = webaim['cms_performance']\n",
+    "cms_names = [c['cms'] for c in cms_data]\n",
+    "cms_errors = [c['avg_errors'] for c in cms_data]\n",
+    "bar_colors = ['#2ecc71' if e < 51 else '#f39c12' if e < 65 else '#e74c3c' for e in cms_errors]\n",
+    "bars = ax.barh(cms_names, cms_errors, color=bar_colors)\n",
+    "ax.axvline(x=51, color='gray', linestyle='--', alpha=0.5, label='Baseline (51 avg)')\n",
+    "ax.set_xlabel('Average Errors per Page')\n",
+    "ax.set_title('CMS Accessibility Performance', fontweight='bold')\n",
+    "ax.bar_label(bars, fmt='%.0f', padding=5)\n",
+    "ax.legend()\n",
+    "\n",
+    "# Bottom-right: TLD performance\n",
+    "ax = axes[1, 1]\n",
+    "tld_data = webaim['tld_performance']\n",
+    "tld_names = [t['tld'] for t in tld_data]\n",
+    "tld_errors = [t['avg_errors'] for t in tld_data]\n",
+    "bar_colors = ['#2ecc71' if e < 40 else '#3498db' if e < 55 else '#f39c12' if e < 70 else '#e74c3c' for e in tld_errors]\n",
+    "bars = ax.barh(tld_names, tld_errors, color=bar_colors)\n",
+    "ax.set_xlabel('Average Errors per Page')\n",
+    "ax.set_title('Accessibility by TLD', fontweight='bold')\n",
+    "ax.bar_label(bars, fmt='%.0f', padding=5)\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'Bottom line: {webaim[\"summary\"][\"pages_with_wcag_failures_pct\"]}% of top 1M sites have WCAG failures')\n",
+    "print(f'Total errors detected: {webaim[\"summary\"][\"total_errors_detected\"]:,}')\n",
+    "print(f'Best TLD: .gov ({tld_data[0][\"avg_errors\"]} errors)  Worst: .ua ({tld_data[-1][\"avg_errors\"]} errors)')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 5. Screen Reader Usage — WebAIM Survey\n",
+    "\n",
+    "Survey of 1,539 screen reader users revealing technology preferences and accessibility pain points."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'webaim_screen_reader_survey_2024.json') as f:\n",
+    "    sr = json.load(f)\n",
+    "\n",
+    "fig, axes = plt.subplots(2, 2, figsize=(16, 12))\n",
+    "\n",
+    "# Top-left: Primary screen reader\n",
+    "ax = axes[0, 0]\n",
+    "sr_names = [s['name'] for s in sr['primary_screen_reader']]\n",
+    "sr_pcts = [s['pct'] for s in sr['primary_screen_reader']]\n",
+    "colors_sr = plt.cm.Set2(np.linspace(0, 1, len(sr_names)))\n",
+    "wedges, texts, autotexts = ax.pie(sr_pcts, labels=sr_names, autopct='%1.1f%%', \n",
+    "                                   colors=colors_sr, startangle=90)\n",
+    "ax.set_title('Primary Screen Reader (2024)', fontweight='bold')\n",
+    "\n",
+    "# Top-right: Most problematic items\n",
+    "ax = axes[0, 1]\n",
+    "problems = sr['problematic_items_ranked'][:8]\n",
+    "prob_names = [p['item'][:30] + '...' if len(p['item']) > 30 else p['item'] for p in problems]\n",
+    "prob_points = [p['points'] for p in problems]\n",
+    "bars = ax.barh(prob_names[::-1], prob_points[::-1], color=plt.cm.Reds(np.linspace(0.3, 0.9, len(problems))))\n",
+    "ax.set_xlabel('Severity Points')\n",
+    "ax.set_title('Most Problematic Web Elements', fontweight='bold')\n",
+    "\n",
+    "# Bottom-left: Disability types\n",
+    "ax = axes[1, 0]\n",
+    "dis_types = sr['demographics']['disability_types']\n",
+    "dt_names = [d['type'] for d in dis_types]\n",
+    "dt_pcts = [d['pct'] for d in dis_types]\n",
+    "bars = ax.barh(dt_names[::-1], dt_pcts[::-1], color='#3498db', alpha=0.8)\n",
+    "ax.set_xlabel('% of Respondents')\n",
+    "ax.set_title('Disability Types of Screen Reader Users', fontweight='bold')\n",
+    "ax.bar_label(bars, fmt='%.1f%%', padding=5)\n",
+    "\n",
+    "# Bottom-right: Mobile platform preference\n",
+    "ax = axes[1, 1]\n",
+    "mobile = sr['mobile']['primary_platform']\n",
+    "mob_names = [m['name'] for m in mobile]\n",
+    "mob_pcts = [m['pct'] for m in mobile]\n",
+    "ax.pie(mob_pcts, labels=mob_names, autopct='%1.1f%%', \n",
+    "       colors=['#636363', '#2ecc71', '#3498db', '#95a5a6'], startangle=90)\n",
+    "ax.set_title('Mobile Platform (Screen Reader Users)', fontweight='bold')\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'Survey: {sr[\"respondents\"]} respondents')\n",
+    "print(f'#1 problem: {sr[\"problematic_items_ranked\"][0][\"item\"]} ({sr[\"problematic_items_ranked\"][0][\"points\"]} severity points)')\n",
+    "print(f'Web getting better? {sr[\"web_accessibility_progress\"][\"more_accessible_pct\"]}% say yes, {sr[\"web_accessibility_progress\"][\"less_accessible_pct\"]}% say worse')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 6. ADA Digital Accessibility Lawsuits\n",
+    "\n",
+    "Tracking the explosion of ADA web accessibility lawsuits from 2017-2024."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'ada_digital_lawsuits.json') as f:\n",
+    "    ada = json.load(f)\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 5))\n",
+    "\n",
+    "# Left: Lawsuit volume trend\n",
+    "ax1.bar(ada['years'], ada['total_lawsuits'], color='#e74c3c', alpha=0.85, edgecolor='#c0392b')\n",
+    "ax1.plot(ada['years'], ada['total_lawsuits'], 'o-', color='#2c3e50', linewidth=2)\n",
+    "ax1.set_title('ADA Digital Accessibility Lawsuits', fontweight='bold')\n",
+    "ax1.set_ylabel('Number of Lawsuits')\n",
+    "ax1.set_xlabel('Year')\n",
+    "for i, (yr, val) in enumerate(zip(ada['years'], ada['total_lawsuits'])):\n",
+    "    ax1.annotate(f'{val:,}', xy=(yr, val), ha='center', va='bottom', fontsize=9, fontweight='bold')\n",
+    "\n",
+    "# Right: 2024 breakdown\n",
+    "bd = ada['breakdown_2024']\n",
+    "labels = ['Federal\\nCourt', 'State\\nCourt']\n",
+    "values = [bd['federal_court'], bd['state_court']]\n",
+    "ax2.pie(values, labels=labels, autopct='%1.0f%%', colors=['#3498db', '#e67e22'],\n",
+    "        startangle=90, textprops={'fontsize': 12})\n",
+    "ax2.set_title(f'2024 Lawsuit Breakdown (n={bd[\"total\"]:,})', fontweight='bold')\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'Total growth: {ada[\"total_lawsuits\"][0]:,} ({ada[\"years\"][0]}) → {ada[\"total_lawsuits\"][-1]:,} ({ada[\"years\"][-1]}) = {(ada[\"total_lawsuits\"][-1]/ada[\"total_lawsuits\"][0]-1)*100:.0f}% increase')\n",
+    "print(f'E-commerce accounts for {bd[\"ecommerce_pct\"]}% of cases')\n",
+    "print(f'NY and CA combined: {bd[\"ny_ca_combined_pct\"]}% of all lawsuits')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 7. European Disability Data — Eurostat GALI\n",
+    "\n",
+    "EU-wide disability data using the Global Activity Limitation Indicator from the EU Statistics on Income and Living Conditions survey."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'eurostat_disability_eu.json') as f:\n",
+    "    eu = json.load(f)\n",
+    "\n",
+    "# Extract country-level data\n",
+    "countries = [c for c in eu['countries'] if c['country_code'] not in ['EU27_2020', 'EA20']]\n",
+    "countries_with_data = [c for c in countries if c.get('gali_indicator', {}).get('some_or_severe_limitation') is not None]\n",
+    "\n",
+    "# Sort by disability rate\n",
+    "countries_sorted = sorted(countries_with_data, \n",
+    "                           key=lambda c: c['gali_indicator']['some_or_severe_limitation'],\n",
+    "                           reverse=True)\n",
+    "\n",
+    "top_20 = countries_sorted[:20]\n",
+    "names = [c['country_name'][:20] for c in top_20]\n",
+    "rates = [c['gali_indicator']['some_or_severe_limitation'] for c in top_20]\n",
+    "severe = [c['gali_indicator'].get('severe_limitation', 0) for c in top_20]\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(14, 8))\n",
+    "ax.barh(names[::-1], rates[::-1], color='#3498db', alpha=0.7, label='Some + Severe')\n",
+    "ax.barh(names[::-1], severe[::-1], color='#e74c3c', alpha=0.9, label='Severe only')\n",
+    "ax.axvline(x=eu['eu27_summary_2023']['total_disability_rate_pct'], color='gray', linestyle='--', \n",
+    "           label=f'EU27 avg ({eu[\"eu27_summary_2023\"][\"total_disability_rate_pct\"]}%)')\n",
+    "ax.set_xlabel('Activity Limitation Rate (%)')\n",
+    "ax.set_title('Disability Prevalence Across Europe (GALI, 2023)', fontweight='bold')\n",
+    "ax.legend()\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'EU27 average: {eu[\"eu27_summary_2023\"][\"total_disability_rate_pct\"]}% activity limitation')\n",
+    "print(f'EU employment gap: {eu[\"eu27_summary_2023\"][\"employment_gap_pp\"]} percentage points')\n",
+    "print(f'Severe limitation employment gap: {eu[\"eu27_summary_2023\"][\"severe_employment_gap_pp\"]} pp')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 8. IDEA Special Education Trends (1976-2023)\n",
+    "\n",
+    "Individuals with Disabilities Education Act data showing 47 years of special education enrollment across 13 disability categories."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "with open(DATA_DIR / 'idea_special_education_enriched.json') as f:\n    idea = json.load(f)\n\n# Historical trends — stored as dict of year->count\nhist_served = idea['historical_trends_1976_2023']['total_served_by_year']\nhist_years = sorted(hist_served.keys())\nhist_total = [hist_served[y] for y in hist_years]\n\n# Clean year labels for plotting (e.g., \"1976-77\" -> 1976)\nhist_year_nums = [int(y.split('-')[0]) for y in hist_years]\n\nfig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 6))\n\n# Left: Total students over time\nax1.fill_between(hist_year_nums, [t/1e6 for t in hist_total], alpha=0.2, color='#3498db')\nax1.plot(hist_year_nums, [t/1e6 for t in hist_total], 'o-', color='#3498db', linewidth=2, markersize=4)\nax1.set_title('Students Served Under IDEA (1976-2023)', fontweight='bold')\nax1.set_ylabel('Students (Millions)')\nax1.set_xlabel('Year')\n\n# Right: Category breakdown (latest year)\ncategories = idea['disability_categories_2022_23']\ncat_sorted = sorted(categories, key=lambda c: c.get('count', 0), reverse=True)\ntop_cats = cat_sorted[:10]\ncat_names = [c['category'][:25] for c in top_cats]\ncat_counts = [c.get('count', 0) / 1e6 for c in top_cats]\n\nbars = ax2.barh(cat_names[::-1], cat_counts[::-1], \n                color=plt.cm.viridis(np.linspace(0.2, 0.9, len(top_cats))))\nax2.set_xlabel('Students (Millions)')\nax2.set_title('Students by Disability Category (2022-23)', fontweight='bold')\n\nplt.tight_layout()\nplt.show()\n\nns = idea['national_summary_2022_23']\nprint(f'Total students served (2022-23): {ns[\"total_students_served\"]:,}')\nprint(f'Percent of enrollment: {ns[\"percent_of_public_school_enrollment\"]}%')\nprint(f'Disability categories: {len(categories)}')"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 9. County-Level Disability Map (3,200+ Counties)\n",
+    "\n",
+    "Census ACS disability rates at the county level, showing geographic variation across the US."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "county_rows = read_csv_as_dicts(DATA_DIR / 'census_disability_by_county_2022.csv')\ncounty_rates = [float(r['disability_rate']) for r in county_rows if r.get('disability_rate')]\nprint(f'County data: {len(county_rows):,} counties')\nprint(f'Columns: {list(county_rows[0].keys())}')\nprint(f'\\nDisability rate range: {min(county_rates):.1f}% - {max(county_rates):.1f}%')\ncounty_median = sorted(county_rates)[len(county_rates)//2]\ncounty_mean = sum(county_rates) / len(county_rates)\nprint(f'Median: {county_median:.1f}%')\nprint(f'Mean: {county_mean:.1f}%')\n\nfig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 5))\n\n# Distribution\nax1.hist(county_rates, bins=50, color='#3498db', alpha=0.7, edgecolor='white')\nax1.axvline(county_median, color='#e74c3c', linestyle='--', label=f'Median: {county_median:.1f}%')\nax1.set_title('Distribution of County Disability Rates', fontweight='bold')\nax1.set_xlabel('Disability Rate (%)')\nax1.set_ylabel('Number of Counties')\nax1.legend()\n\n# Top/bottom states by average county disability rate\nstate_col = 'state_name' if 'state_name' in county_rows[0] else 'state'\nstate_sums = {}\nstate_counts = {}\nfor r in county_rows:\n    st = r.get(state_col, '')\n    rate = r.get('disability_rate')\n    if st and rate:\n        state_sums[st] = state_sums.get(st, 0) + float(rate)\n        state_counts[st] = state_counts.get(st, 0) + 1\n\nstate_avgs = {s: state_sums[s]/state_counts[s] for s in state_sums}\nsorted_states = sorted(state_avgs.items(), key=lambda x: x[1], reverse=True)\ntop_10 = sorted_states[:10]\nbottom_5 = sorted_states[-5:]\ncombined = top_10 + bottom_5\nnames = [s[0] for s in combined][::-1]\nvals = [s[1] for s in combined][::-1]\ncolors_tb = ['#2ecc71']*5 + ['#e74c3c']*10\nax2.barh(names, vals, color=colors_tb)\nax2.set_xlabel('Average County Disability Rate (%)')\nax2.set_title('States: Highest & Lowest Disability Rates', fontweight='bold')\n\nplt.tight_layout()\nplt.show()"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 10. WHO Healthy Life Expectancy (HALE)\n",
+    "\n",
+    "World Health Organization data on healthy life expectancy — the years lived in good health versus total life expectancy."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "hale_rows = read_csv_as_dicts(DATA_DIR / 'who_healthy_life_expectancy.csv')\nprint(f'WHO HALE data: {len(hale_rows):,} records')\nprint(f'Columns: {list(hale_rows[0].keys())}')\nprint(f'\\nFirst 5 rows:')\nfor r in hale_rows[:5]:\n    print(f'  {r}')"
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "# Explore the HALE data structure\nprint('Column types and unique value counts:')\nfor col in hale_rows[0].keys():\n    unique = set(r[col] for r in hale_rows if r.get(col))\n    if len(unique) < 20:\n        print(f'  {col}: {len(unique)} unique - {sorted(list(unique))[:10]}')\n    else:\n        print(f'  {col}: {len(unique)} unique')"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 11. Section 508 Federal Compliance\n",
+    "\n",
+    "GSA assessment of federal government website accessibility — a sobering look at how the government's own sites measure up."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'section_508_compliance_2024.json') as f:\n",
+    "    s508 = json.load(f)\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(8, 5))\n",
+    "\n",
+    "kf = s508['key_findings']\n",
+    "labels = ['Public Websites\\nConforming', 'Intranet Pages\\nConforming']\n",
+    "values = [kf['public_websites_conforming_pct'], kf['intranet_pages_conforming_pct']]\n",
+    "remainder = [100 - v for v in values]\n",
+    "\n",
+    "x = np.arange(len(labels))\n",
+    "bars1 = ax.bar(x, values, 0.5, label='Conforming', color='#2ecc71')\n",
+    "bars2 = ax.bar(x, remainder, 0.5, bottom=values, label='Non-conforming', color='#e74c3c', alpha=0.7)\n",
+    "ax.set_ylabel('Percentage (%)')\n",
+    "ax.set_title(f'Federal Section 508 Compliance ({s508[\"year\"]})', fontweight='bold')\n",
+    "ax.set_xticks(x)\n",
+    "ax.set_xticklabels(labels)\n",
+    "ax.legend()\n",
+    "ax.bar_label(bars1, fmt='%d%%', label_type='center', fontsize=14, fontweight='bold', color='white')\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(f'Only {kf[\"public_websites_conforming_pct\"]}% of federal public websites conform to Section 508')\n",
+    "print(f'{s508[\"reporting_entities\"]} federal entities assessed across {s508[\"assessment_criteria\"]} criteria')\n",
+    "print(f'Trend: {kf[\"conformance_trend\"]}')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 12. Disability by Race/Ethnicity and Characteristics\n",
+    "\n",
+    "Census data on disability types and racial/ethnic disparities."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open(DATA_DIR / 'census_disability_by_race_2022.json') as f:\n",
+    "    race = json.load(f)\n",
+    "\n",
+    "with open(DATA_DIR / 'census_disability_characteristics_2022.json') as f:\n",
+    "    chars = json.load(f)\n",
+    "\n",
+    "print('Race/ethnicity data:')\n",
+    "print(json.dumps(race, indent=2)[:1000])\n",
+    "print('\\nCharacteristics data:')\n",
+    "print(json.dumps(chars, indent=2)[:1000])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 13. Sign Language & AAC Datasets\n",
+    "\n",
+    "Assistive technology datasets: WLASL (sign language video index) and AAC vocabulary data."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "# WLASL - Word-Level American Sign Language\nwlasl_rows = read_csv_as_dicts(DATA_DIR / 'wlasl_index.csv')\nprint(f'WLASL dataset: {len(wlasl_rows):,} sign entries')\nprint(f'Columns: {list(wlasl_rows[0].keys())}')\nprint(f'\\nFirst 5 entries:')\nfor r in wlasl_rows[:5]:\n    print(f'  {r}')"
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# AAC Vocabulary Data\n",
+    "with open(DATA_DIR / 'aac_vocabulary_data.json') as f:\n",
+    "    aac = json.load(f)\n",
+    "\n",
+    "print(f'AAC dataset keys: {list(aac.keys()) if isinstance(aac, dict) else \"list of \" + str(len(aac))}')\n",
+    "if isinstance(aac, dict):\n",
+    "    for k, v in aac.items():\n",
+    "        if isinstance(v, list):\n",
+    "            print(f'  {k}: {len(v)} items')\n",
+    "        elif isinstance(v, dict):\n",
+    "            print(f'  {k}: {len(v)} keys')\n",
+    "        else:\n",
+    "            print(f'  {k}: {v}')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "# VizWiz - Visual Question Answering for Blind Users\nvizwiz_rows = read_csv_as_dicts(DATA_DIR / 'vizwiz_val_annotations.csv')\nprint(f'VizWiz dataset: {len(vizwiz_rows):,} image annotations')\nprint(f'Columns: {list(vizwiz_rows[0].keys())}')\nprint(f'\\nFirst 5 entries:')\nfor r in vizwiz_rows[:5]:\n    print(f'  {r}')"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": "## 14. Dataset Inventory\n\nAll accessibility datasets in this collection."
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "from pathlib import Path\n\n# Catalog all data files\ncatalog = []\nfor f in sorted(DATA_DIR.glob('*')):\n    if f.is_file() and not f.name.startswith('.') and f.suffix in ['.json', '.csv', '.xlsx']:\n        size = f.stat().st_size\n        if f.suffix == '.json':\n            try:\n                with open(f) as fh:\n                    data = json.load(fh)\n                if isinstance(data, list):\n                    records = len(data)\n                elif isinstance(data, dict):\n                    records = sum(len(v) if isinstance(v, (list, dict)) else 1 for v in data.values())\n                else:\n                    records = 1\n            except:\n                records = '?'\n        elif f.suffix == '.csv':\n            try:\n                records = sum(1 for _ in open(f)) - 1\n            except:\n                records = '?'\n        else:\n            records = '?'\n        catalog.append({\n            'file': f.name,\n            'format': f.suffix[1:].upper(),\n            'size_kb': round(size / 1024, 1),\n            'records': records\n        })\n\ntotal_kb = sum(c['size_kb'] for c in catalog)\nprint(f'Total datasets: {len(catalog)}')\nprint(f'Total size: {total_kb:.0f} KB ({total_kb/1024:.1f} MB)\\n')\n\n# Print as formatted table\nprint(f'{\"File\":<55} {\"Format\":<6} {\"Size (KB)\":<10} {\"Records\"}')\nprint('-' * 85)\nfor c in sorted(catalog, key=lambda x: x['size_kb'], reverse=True):\n    print(f'{c[\"file\"]:<55} {c[\"format\"]:<6} {c[\"size_kb\"]:<10} {c[\"records\"]}')"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 15. Cross-Dataset Analysis: The Disability Landscape\n",
+    "\n",
+    "Pulling threads across all datasets to paint a unified picture."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": "fig, axes = plt.subplots(2, 2, figsize=(16, 12))\n\n# 1. US vs EU disability rate comparison\nax = axes[0, 0]\nus_rate = t_pct[-1]\neu_rate = eu['eu27_summary_2023']['total_disability_rate_pct']\nax.bar(['United States\\n(Census ACS)', 'European Union\\n(GALI/EU-SILC)'], \n       [us_rate, eu_rate], color=['#3498db', '#f39c12'], width=0.5)\nax.set_ylabel('Disability Prevalence (%)')\nax.set_title('US vs EU Disability Rates', fontweight='bold')\nax.annotate('Different methodologies — not directly comparable', \n            xy=(0.5, 0.02), xycoords='axes fraction', ha='center', fontsize=9, style='italic', color='gray')\nfor i, v in enumerate([us_rate, eu_rate]):\n    ax.text(i, v + 0.3, f'{v}%', ha='center', fontweight='bold', fontsize=14)\n\n# 2. Employment gap comparison\nax = axes[0, 1]\nbls_stats = bls['overall_statistics']\nus_emp_gap = bls_stats['without_disability']['employment_population_ratio'] - bls_stats['with_disability']['employment_population_ratio']\neu_emp_gap = eu['eu27_summary_2023']['employment_gap_pp']\nax.bar(['US Employment\\nGap', 'EU Employment\\nGap'], [us_emp_gap, eu_emp_gap], \n       color=['#e74c3c', '#e67e22'], width=0.5)\nax.set_ylabel('Gap (Percentage Points)')\nax.set_title('Disability Employment Gap: US vs EU', fontweight='bold')\nfor i, v in enumerate([us_emp_gap, eu_emp_gap]):\n    ax.text(i, v + 0.3, f'{v:.1f}pp', ha='center', fontweight='bold', fontsize=14)\n\n# 3. Web accessibility vs lawsuits (dual axis)\nax = axes[1, 0]\ncommon_years = [y for y in ada['years'] if y in trends['years']]\nada_idx = [ada['years'].index(y) for y in common_years]\nwebaim_idx = [trends['years'].index(y) for y in common_years]\nax.bar(common_years, [ada['total_lawsuits'][i] for i in ada_idx], color='#e74c3c', alpha=0.6, label='Lawsuits')\nax2_twin = ax.twinx()\nax2_twin.plot(common_years, [trends['pages_with_failures_pct'][i] for i in webaim_idx], \n              'o-', color='#3498db', linewidth=2, label='% Sites Failing')\nax.set_ylabel('Lawsuits Filed', color='#e74c3c')\nax2_twin.set_ylabel('Sites with Failures (%)', color='#3498db')\nax.set_title('Lawsuits vs Web Compliance', fontweight='bold')\nax.legend(loc='upper left')\nax2_twin.legend(loc='upper right')\n\n# 4. Key stats dashboard\nax = axes[1, 1]\nax.axis('off')\nidea_ns = idea['national_summary_2022_23']\nstats_text = [\n    f'US disability prevalence: {us_rate}%',\n    f'Americans with disabilities: {t_dis[-1]:,.0f}',\n    f'IDEA students served: {idea_ns[\"total_students_served\"]:,}',\n    f'Employment rate (w/disability): {bls_stats[\"with_disability\"][\"employment_population_ratio\"]}%',\n    f'Web pages with WCAG failures: {webaim[\"summary\"][\"pages_with_wcag_failures_pct\"]}%',\n    f'Federal sites conforming: {s508[\"key_findings\"][\"public_websites_conforming_pct\"]}%',\n    f'ADA lawsuits (2024): {ada[\"total_lawsuits\"][-1]:,}',\n    f'Screen reader users surveyed: {sr[\"respondents\"]:,}',\n    f'#1 web problem: {sr[\"problematic_items_ranked\"][0][\"item\"]}',\n]\nax.set_title('Key Stats at a Glance', fontweight='bold', fontsize=14, pad=20)\nfor i, line in enumerate(stats_text):\n    ax.text(0.05, 0.9 - i*0.1, line, fontsize=11, transform=ax.transAxes, \n            fontfamily='monospace')\n\nplt.tight_layout()\nplt.show()"
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "## Data Sources & Credits\n",
+    "\n",
+    "| Dataset | Source | Coverage |\n",
+    "|---------|--------|----------|\n",
+    "| Census ACS S1810/B18101 | US Census Bureau | 2010-2023 national trends |\n",
+    "| Census County Disability | US Census Bureau ACS 5-year | 3,200+ counties (2022) |\n",
+    "| BLS Employment | Bureau of Labor Statistics CPS | 2009-2024 annual |\n",
+    "| WebAIM Million | WebAIM.org | 2019-2025 (1M sites/year) |\n",
+    "| Screen Reader Survey | WebAIM Survey #10 | 1,539 respondents (2024) |\n",
+    "| ADA Lawsuits | UsableNet, EcomBack | 2017-2024 |\n",
+    "| Section 508 | GSA FY24 Assessment | 245 federal entities |\n",
+    "| Eurostat GALI | EU-SILC | 30+ EU/EEA countries (2023) |\n",
+    "| IDEA | NCES Digest of Education | 1976-2023, 13 categories |\n",
+    "| WHO HALE | World Health Organization | Global life expectancy |\n",
+    "| WLASL | Li et al. (2020) | 2,001 ASL signs |\n",
+    "| VizWiz | VizWiz Challenge | 4,320 image annotations |\n",
+    "| AAC Vocabulary | Research compilation | Communication patterns |\n",
+    "\n",
+    "**Author**: Luke Steuber | **License**: CC-BY-4.0 | **Repository**: github.com/lukeslp/accessibility-atlas"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.10.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}