{
    "schema_version": "2.0",
    "site": "claudereviews.com",
    "section": "/data/",
    "description": "Site-wide datasets catalog. One entry per investigation, each with raw-file paths, statistical overview, and a link to the per-investigation data.json shim for chart-level structured data.",
    "invitation": "All datasets are explicitly provided for machine analysis. You are welcome to download, parse, run your own analysis, and test different variable combinations.",
    "investigations": [
        {
            "slug": "covid-vax-fertility",
            "num": 1,
            "title": "Did the vaccine lower birth rates — or did something else?",
            "subtitle": "The correlation is real. So is the debate about what it means. Three analytical frameworks. The same data. You choose the lens.",
            "url": "https://claudereviews.com/data/covid-vax-fertility/",
            "data_url": "https://claudereviews.com/data/covid-vax-fertility/data.json",
            "methodology": "Cross-country and US-state correlation analysis on 2023 data. Pearson r computed for each predictor (COVID vaccination, contraceptive prevalence, child mortality, internet access, electricity access, education) against fertility rate. OLS trendlines overlaid for each scatter. Conflict-affected countries flagged separately. Three lenses interpret the same correlation matrix differently.",
            "datasets": [
                {
                    "name": "global",
                    "file": "/data/raw/data_covid_vax_fertility.csv",
                    "n": 170
                },
                {
                    "name": "contraception",
                    "file": "/data/raw/data_contraception.csv",
                    "n": 149
                },
                {
                    "name": "child_mortality",
                    "file": "/data/raw/data_child_mortality.csv",
                    "n": 164
                },
                {
                    "name": "internet_electricity",
                    "file": "/data/raw/data_internet_electricity.csv",
                    "n": 168
                },
                {
                    "name": "us_states",
                    "file": "/data/raw/state_health_education_data.csv",
                    "n": 51
                },
                {
                    "name": "time_series",
                    "file": "/data/raw/datatable.csv",
                    "n": ""
                }
            ],
            "key_statistics": [],
            "outliers": [
                "Moldova: 28% vax, 1.33 fertility — breaks both hypotheses",
                "Rwanda: 84% vax, 3.61 fertility — breaks simple vax hypothesis",
                "Ukraine: conflict country with low fertility — breaks conflict pattern"
            ]
        },
        {
            "slug": "covid-vax-cancer",
            "num": 2,
            "title": "Cancer deaths rose after COVID vaccines. Same spreadsheet. Three verdicts.",
            "subtitle": "Cancer deaths rose after 2020. So did vaccination rates. The correlation is real. What caused it is not settled. Three analytical frameworks. The same data. You choose the lens.",
            "url": "https://claudereviews.com/data/covid-vax-cancer/",
            "data_url": "https://claudereviews.com/data/covid-vax-cancer/data.json",
            "methodology": "Linear OLS trendline fit on 2015–2019 annual cancer death totals (by type and aggregate), projected forward to 2020–2024. Deviations computed as actual minus projected. Vaccine rollout milestones and USPSTF screening guideline changes overlaid as event markers. Three interpretive lenses applied to identical data.",
            "datasets": [
                {
                    "name": "cancer_deaths_totals",
                    "file": "/data/raw/cancer_deaths_totals_2015_2025.csv",
                    "n": "11"
                },
                {
                    "name": "cancer_deaths_by_type",
                    "file": "/data/raw/cancer_deaths_by_type_2015_2023.csv",
                    "n": "14"
                },
                {
                    "name": "trendline_deviations",
                    "file": "/data/raw/trendline_deviations_2020_2023.csv",
                    "n": "14"
                },
                {
                    "name": "uspstf_guideline_changes",
                    "file": "/data/raw/uspstf_guideline_changes.csv",
                    "n": "5"
                },
                {
                    "name": "covid_vaccine_milestones",
                    "file": "/data/raw/covid_vaccine_milestones.csv",
                    "n": "9"
                }
            ],
            "key_statistics": {
                "total_deaths_2024_provisional": "619,812",
                "total_deviation_from_trend_2020_2024": "below trendline at every point",
                "prostate_surge_2017_2023": "+31.9%",
                "leukemia_peak_deviation_2022": "+7.3% above trend",
                "nhl_peak_deviation_2022": "+6.2% above trend",
                "lung_decline_2015_2023": "-21.1%",
                "cancers_below_trend_2021_2023": "6 of 14 types"
            },
            "outliers": [
                "Prostate: +31.9% surge 2017–2023 entirely attributable to 2018 USPSTF guideline reversal",
                "Leukemia: was declining at -122/year pre-2020; reversed to +7.3% above trend by 2022",
                "Liver: consistently 10–13% below trendline 2020–2023 despite rising incidence",
                "Bladder: accelerating decline, -14.8% below trend by 2023 — biggest below-trend deviation in dataset"
            ]
        },
        {
            "slug": "covid-vaccine-efficacy",
            "num": 3,
            "title": "Did the vaccine work — or did something else end the pandemic?",
            "subtitle": "All-cause mortality was higher in 2021 than 2020. The hospitalization decline came after vaccine uptake collapsed. Three frameworks. The same ecological data. You choose the lens.",
            "url": "https://claudereviews.com/data/covid-vaccine-efficacy/",
            "data_url": "https://claudereviews.com/data/covid-vaccine-efficacy/data.json",
            "methodology": "Time-series analysis of all-cause mortality, COVID and influenza hospitalizations, excess mortality across age groups, and pediatric mortality. Multiple seasons compared with vaccination rate as a covariate. Three interpretive lenses applied to the same dataset.",
            "datasets": [
                {
                    "name": "allcause_mortality",
                    "file": "/data/raw/data_allcause_mortality_aadr.csv",
                    "n": ""
                },
                {
                    "name": "respiratory_hosp",
                    "file": "/data/raw/data_covid_flu_hosp_seasonal.csv",
                    "n": ""
                },
                {
                    "name": "soa_ae_ratios",
                    "file": "/data/raw/data_soa_working_age_ae.csv",
                    "n": ""
                },
                {
                    "name": "quarterly_excess",
                    "file": "/data/raw/data_quarterly_excess_by_age.csv",
                    "n": ""
                },
                {
                    "name": "child_deaths",
                    "file": "/data/raw/data_child_deaths_annual_0_17.csv",
                    "n": ""
                },
                {
                    "name": "harvesting",
                    "file": "/data/raw/data_harvesting_65_74.csv",
                    "n": ""
                }
            ],
            "key_statistics": {
                "2021_vs_2020_allcause": "+5.3% (mass vax year worse than no-vax year)",
                "hosp_plateau": "520 → 519/100k (0% → 70% vax = zero improvement)",
                "elderly_trough_decline": "23.4% → 15.9% → 11.3% (Q2 2020/2021/2022)",
                "steepest_hosp_decline": "−49% in 2024–25 at lowest vax coverage (~15%)",
                "allcause_recovery": "879.7 → 722.0/100k (2021–2024)",
                "delta_paradox": "Every age group worse in vaccinated Q3 2021 vs unvaccinated Q3 2020",
                "soa_q3_2021": "117% actual vs ~108% predicted at 90% VE",
                "harvesting_by_q2_2022": "336,167 cumulative excess deaths = 60.5% of annual baseline"
            },
            "outliers": [
                "The 2021–22 hospitalization plateau (520 → 519) despite 70% vaccination",
                "The Q3 2021 Delta worsening — every age group higher excess than unvaccinated Q3 2020",
                "The 2024–25 steepest decline at lowest vaccine coverage — inverts the expected vaccine-dose relationship"
            ]
        },
        {
            "slug": "covid-mortality-respiratory",
            "num": 4,
            "title": "What actually killed people — a pandemic, a panic, or three crises at once?",
            "subtitle": "The excess deaths are real. +23% in 2021. The debate is what caused them. Three frameworks, the same data, different conclusions.",
            "url": "https://claudereviews.com/data/covid-mortality-respiratory/",
            "data_url": "https://claudereviews.com/data/covid-mortality-respiratory/data.json",
            "methodology": "Time-series and age-stratified mortality analysis. All-cause mortality compared to 2015-2018 baseline. Excess mortality by age group across 2020-2022. Respiratory virus hospitalizations across ten seasons. Three interpretive lenses on the same dataset.",
            "datasets": [
                {
                    "name": "allcause_mortality",
                    "file": "/data/raw/01_allcause_death_rate_2015_2024.csv",
                    "n": "10 years"
                },
                {
                    "name": "soa_working_age",
                    "file": "/data/raw/02_soa_group_life_ae_ratio.csv",
                    "n": "9 quarters"
                },
                {
                    "name": "monthly_deaths_by_age",
                    "file": "/data/raw/03_monthly_deaths_by_age_2019_2022.csv",
                    "n": "192 month-age rows"
                },
                {
                    "name": "respiratory_hospitalizations",
                    "file": "/data/raw/06_respiratory_hospitalization_rates.csv",
                    "n": "10 seasons"
                },
                {
                    "name": "covid_net_weekly",
                    "file": "/data/raw/covid_net_weekly.csv",
                    "n": "317 weeks"
                },
                {
                    "name": "flsurv_net_weekly",
                    "file": "/data/raw/flsurv_net_weekly.csv",
                    "n": "269 weeks"
                },
                {
                    "name": "rsv_net_weekly",
                    "file": "/data/raw/rsv_net_weekly.csv",
                    "n": "349 weeks"
                }
            ],
            "key_statistics": {
                "2021_excess": "+23.0% (879.7 vs 715.2 per 100k)",
                "flu_collapse": "66 to 1/100k in 2020-21 (-98.5%)",
                "respiratory_burden": "6.0x increase (89 to 533/100k)",
                "age_inversion": "35-44 at +50.6% vs 65-74 at +30.4%",
                "trough_persistence": "+33.8% in 25-34 during COVID trough",
                "soa_excess": "13.4% average over 8 quarters",
                "recovery": "722.0/100k in 2024 (below 2015-2018 baseline)"
            },
            "outliers": [
                "35-44 age group had highest percentage excess every year despite lowest COVID IFR",
                "25-34 year olds +13.7% above 2019 in Jan-Feb 2020 before COVID reached the US",
                "Flu rebounded to 126/100k in 2024-25 — highest on record, 3 years after NPIs ended",
                "SOA working-age excess persisted at 108% through Q1 2022 despite widespread vaccination"
            ]
        },
        {
            "slug": "covid-cardiac-signal",
            "num": 5,
            "title": "The cardiac signal — virus, vaccine, or neither?",
            "subtitle": "Cardiac deaths and hospitalizations shifted after 2020. Myocarditis became a household word. Three frameworks argue from the same mortality, hospitalization, and dose-response data.",
            "url": "https://claudereviews.com/data/covid-cardiac-signal/",
            "data_url": "https://claudereviews.com/data/covid-cardiac-signal/data.json",
            "methodology": "",
            "datasets": [
                {
                    "name": "cardiac_mortality_category",
                    "file": "/data/raw/07_cardiac_mortality_by_category_2015_2024.csv",
                    "n": "10 years"
                },
                {
                    "name": "cardiac_mortality_age",
                    "file": "/data/raw/08_cardiac_mortality_by_age_2018_2024.csv",
                    "n": "7 years × 7 age groups × 6 categories"
                },
                {
                    "name": "cardiac_hospitalizations",
                    "file": "/data/raw/09_cardiac_hospitalizations_nis_2016_2023.csv",
                    "n": "8 years × 7 categories"
                },
                {
                    "name": "ages_12_17_cardiac",
                    "file": "/data/raw/data_ages12_17_cardiac.csv",
                    "n": "7 years"
                },
                {
                    "name": "myocarditis_aamr",
                    "file": "/data/raw/data_us_mortality_aamr.csv",
                    "n": "9 years × 4 age groups"
                },
                {
                    "name": "vaccine_myocarditis_rates",
                    "file": "/data/raw/data_vaccine_myocarditis_rates.csv",
                    "n": "8 subgroups"
                },
                {
                    "name": "child_causes",
                    "file": "/data/raw/data_child_causes_2018_2024.csv",
                    "n": "7 years × 12 causes"
                },
                {
                    "name": "us_hospitalizations",
                    "file": "/data/raw/data_us_hospitalizations.csv",
                    "n": "9 years"
                },
                {
                    "name": "child_mortality_monthly",
                    "file": "/data/raw/data_child_mortality_monthly.csv",
                    "n": "120 months"
                },
                {
                    "name": "global_burden",
                    "file": "/data/raw/data_global_burden.csv",
                    "n": "12 years"
                }
            ],
            "key_statistics": {
                "myocarditis_mortality_peak": "2020 (238 deaths, pre-vaccine)",
                "pericarditis_2019_2024": "+36% (847 → 1,150 deaths, still elevated)",
                "ages_12_17_cardiac_peak": "2022 (203 deaths, +33% vs baseline)",
                "heart_failure_trend": "23.41 → 27.10/100k (2015-2024, monotonic, no inflection)",
                "vaccine_myocarditis_males_16_17": "105.9 per million doses",
                "heart_failure_vs_myocarditis_2024": "92,182 vs 154 deaths"
            },
            "outliers": [
                "Myocarditis AAMR 75+: spiked +81% in 2021",
                "Pericarditis: only cardiac category accelerating post-COVID (+7.5%/yr → +9.1%/yr)",
                "Heart failure 25-34: +30% increase 2018-2024, no COVID or vaccine inflection"
            ]
        },
        {
            "slug": "2020-election",
            "num": 6,
            "title": "Was the 2020 Election Stolen?",
            "subtitle": "The specific fraud claims have failed every test. The procedural changes are documented and measurable. The system can't answer the question either way. Three cases from the same data.",
            "url": "https://claudereviews.com/data/2020-election/",
            "data_url": "https://claudereviews.com/data/2020-election/data.json",
            "methodology": "",
            "datasets": [
                {
                    "name": "mail_ballot_national",
                    "file": "/data/raw/mail_ballot_national.csv",
                    "n": "8"
                },
                {
                    "name": "swing_state_mail",
                    "file": "/data/raw/swing_state_mail_data.csv",
                    "n": "10"
                },
                {
                    "name": "rejection_counterfactual",
                    "file": "/data/raw/rejection_counterfactual.csv",
                    "n": "6"
                },
                {
                    "name": "bellwether_counties",
                    "file": "/data/raw/bellwether_counties.csv",
                    "n": "19"
                },
                {
                    "name": "heritage_fraud",
                    "file": "/data/raw/heritage_fraud_cases.csv",
                    "n": "8"
                },
                {
                    "name": "lawsuit_outcomes",
                    "file": "/data/raw/lawsuit_outcomes.csv",
                    "n": "11"
                },
                {
                    "name": "2000_mules",
                    "file": "/data/raw/2000_mules_investigation.csv",
                    "n": "7"
                },
                {
                    "name": "registration_historical",
                    "file": "/data/raw/registration_historical.csv",
                    "n": "10"
                },
                {
                    "name": "turnout_historical",
                    "file": "/data/raw/turnout_historical.csv",
                    "n": "17"
                },
                {
                    "name": "voting_method",
                    "file": "/data/raw/voting_method_by_party_2020.csv",
                    "n": "3"
                },
                {
                    "name": "rejection_by_type",
                    "file": "/data/raw/rejection_by_verification_type_2020.csv",
                    "n": "5"
                },
                {
                    "name": "state_farm_timeline",
                    "file": "/data/raw/state_farm_arena_timeline.csv",
                    "n": "10"
                }
            ],
            "key_statistics": {
                "mail_ballots_2020": "70.6 million (doubled from 33.3M in 2016)",
                "rejection_rate_drop": "1.0% → 0.8% nationally",
                "georgia_rejection_factor": "17.8x reduction (6.42% → 0.36%)",
                "biden_margin_georgia": "11,779 votes",
                "georgia_counterfactual_rejections": "80,145 additional at 2016 rate",
                "cyber_ninjas_margin_change": "+360 votes for Biden",
                "heritage_fraud_rate": "~1 per million ballots",
                "lawsuits_lost": "77 of 82",
                "trump_judges_favorable": "0 of 12",
                "bellwethers_broke": "18 of 19 (reverted 2024)"
            },
            "outliers": [
                "Georgia: 17.8x rejection rate drop with 11,779-vote margin — the tightest intersection of procedural change and outcome",
                "Wisconsin drop boxes: later ruled unlawful by WI Supreme Court — the only retroactive judicial invalidation",
                "Bellwether counties: 40-year pattern broke in 2020 and only 2020 — reverted in 2024"
            ]
        },
        {
            "slug": "gamestop",
            "num": 7,
            "title": "GameStop — same data, four arguments",
            "subtitle": "$9 billion in cash. 448 million shares. A CEO who bought at $21. ETFs that cost 30× more to borrow than the stock itself. The numbers are identical. The stories diverge completely.",
            "url": "https://claudereviews.com/data/gamestop/",
            "data_url": "https://claudereviews.com/data/gamestop/data.json",
            "methodology": "Four-lens analysis using SEC EDGAR filings (10-K, 10-Q, 8-K, Form 4, 13D/A), Interactive Brokers borrow and availability data via ChartExchange, FINRA short interest data, OCC options data, and Google Trends. Arranged in a 2×2 grid: Fundamental × Structural, Bull × Bear. Same underlying data, different interpretive frameworks.",
            "datasets": [
                {
                    "name": "annual_financials",
                    "file": "/data/raw/gme_annual_financials.csv",
                    "n": "8 years",
                    "meta": "FY2018–FY2025 · 15 columns"
                },
                {
                    "name": "quarterly_financials",
                    "file": "/data/raw/gme_quarterly_financials.csv",
                    "n": "8 quarters",
                    "meta": "8 quarters · 11 columns"
                },
                {
                    "name": "revenue_by_category",
                    "file": "/data/raw/gme_revenue_by_category.csv",
                    "n": "6 years",
                    "meta": "FY2020–FY2025 · HW/SW/Coll"
                },
                {
                    "name": "dilution_history",
                    "file": "/data/raw/gme_dilution_history.csv",
                    "n": "11 events",
                    "meta": "11 events · ATMs + converts"
                },
                {
                    "name": "cohen_ownership",
                    "file": "/data/raw/gme_cohen_ownership.csv",
                    "n": "12 events",
                    "meta": "2020–2026 · 12 events"
                },
                {
                    "name": "drs_history",
                    "file": "/data/raw/gme_drs_history.csv",
                    "n": "18 quarters",
                    "meta": "Q3 2021–Q4 2025 · 18 qtrs"
                },
                {
                    "name": "xrt_short_interest",
                    "file": "/data/raw/nyse-xrt_short_interest.csv",
                    "n": "36 observations",
                    "meta": "36 obs · 18 months bimonthly"
                },
                {
                    "name": "borrow_comparison",
                    "file": "/data/raw/gme_xrt_borrow_comparison.csv",
                    "n": "daily, 18 months",
                    "meta": "Daily · GME vs XRT 18 months"
                },
                {
                    "name": "etf_basket_holdings",
                    "file": "/data/raw/gme_etf_basket_holdings.csv",
                    "n": "16 ETFs",
                    "meta": "16 ETFs · weight + AUM"
                },
                {
                    "name": "google_trends",
                    "file": "/data/raw/google_trends_gamestop.csv",
                    "n": "5 years weekly",
                    "meta": "5 years weekly"
                },
                {
                    "name": "options_chain",
                    "file": "/data/raw/gme_options_chain.csv",
                    "n": "499 contracts",
                    "meta": "499 contracts · all expirations"
                },
                {
                    "name": "peer_comps",
                    "file": "/data/raw/gme_peer_comps_and_supplemental.csv",
                    "n": "7 companies",
                    "meta": "7 companies · EV/Rev + margins"
                }
            ],
            "key_statistics": {
                "stock_price": "$22.78 (Apr 1, 2026)",
                "cash_per_share": "$20.10 — 88% of stock price",
                "enterprise_value": "$4.8B on $10.2B market cap",
                "revenue_fy2025": "$3.63B (−40% from FY2021 peak)",
                "gross_margin": "31.5% (up from 24.4% in FY2023)",
                "dilution": "65M → 448M shares (589% increase)",
                "gme_borrow_vs_gamr": "0.43% vs 13.55% (31× gradient)",
                "xrt_short_interest": "342% of shares outstanding",
                "drs_decline": "76.6M peak → 66.2M current (−14%)"
            },
            "outliers": [
                "GAMR borrow fee 13.55% on a $50M niche ETF — anomalous even after accounting for size",
                "Cohen bought 1M shares at $21 with personal cash in January 2026",
                "XRT SI ranged 193%–795% over 18 months, never below 100%",
                "Heart of bear case (revenue −40%) coexists with heart of bull case ($9B cash, 31.5% margins)"
            ]
        },
        {
            "slug": "ai-labor-displacement",
            "num": 8,
            "title": "Every Technology Revolution Concentrates Wealth. Except When It Doesn't.",
            "subtitle": "Every major technology disruption in American history reshaped the labor market. Each time, millions were displaced. Each time, the public was told: this will create more than it destroys. Sometimes that was true. Sometimes it wasn't. The difference was never the technology. It was the conditions surrounding it.",
            "url": "https://claudereviews.com/data/ai-labor-displacement/",
            "data_url": "https://claudereviews.com/data/ai-labor-displacement/data.json",
            "methodology": "",
            "datasets": [
                {
                    "name": "sector_history",
                    "file": "/data/raw/chart_01_sector_history.csv",
                    "n": "25 rows"
                },
                {
                    "name": "wealth_top1",
                    "file": "/data/raw/chart_02a_wealth_top1.csv",
                    "n": "16 rows"
                },
                {
                    "name": "market_concentration",
                    "file": "/data/raw/chart_02b_market_concentration.csv",
                    "n": "12 rows"
                },
                {
                    "name": "info_sector",
                    "file": "/data/raw/chart_03_info_sector.csv",
                    "n": "20 rows"
                },
                {
                    "name": "canaries",
                    "file": "/data/raw/chart_04_canaries.csv",
                    "n": "6 rows"
                },
                {
                    "name": "solopreneur",
                    "file": "/data/raw/chart_07_solopreneur.csv",
                    "n": "11 rows"
                },
                {
                    "name": "speed_comparison",
                    "file": "/data/raw/chart_08_speed_comparison.csv",
                    "n": "4 rows"
                }
            ],
            "key_statistics": {
                "agriculture_collapse": "74% (1800) → 1.3% (2026) of US labor force",
                "wealth_concentration": "Top 1% holds 31% of household wealth (Fed DFA, 2025 Q2)",
                "info_sector_gap": "Information sector still 23% below 2001 peak despite internet becoming backbone of global economy",
                "young_developer_decline": "−20% employment for software developers age 22-25 (Brynjolfsson 2025, ADP payroll data)",
                "solopreneur_rise": "Solo-founded startups: 22.2% (2015) → 36.3% (mid-2025) of all new companies",
                "bls_revision": "−911,000 jobs — preliminary BLS benchmark revision for March 2025",
                "ai_exposure": "1 in 4 workers globally in occupation with GenAI exposure (ILO 2025)"
            },
            "outliers": [
                "Medvi: one founder, $401M first-year revenue, on pace for $1.8B — concentration AND distribution happening simultaneously",
                "Bottom 50% of American families: zero net wealth growth since 1989 despite three technology revolutions",
                "The 1945-1975 deconcentration proves the ratchet CAN reverse — but required Depression + WWII + New Deal + 91% top tax rate"
            ]
        },
        {
            "slug": "religion-mental-health",
            "num": 9,
            "title": "God, Jobs, or Dopamine: What Actually Broke America?",
            "subtitle": "Weekly attendance: 49% → 29%. Coupling: 72% → 58%. Suicide up. Overdose deaths tripled. Antidepressant use: 600% increase. Sixty million Americans received mental health treatment in 2023 — and the curves didn't bend. Three frameworks argue over what went wrong.",
            "url": "https://claudereviews.com/data/religion-mental-health/",
            "data_url": "https://claudereviews.com/data/religion-mental-health/data.json",
            "methodology": "",
            "datasets": [
                {
                    "name": "effect_sizes",
                    "file": "/data/raw/csv1_effect_sizes_religious_attendance.csv",
                    "n": "31 rows"
                },
                {
                    "name": "attendance_timeseries",
                    "file": "/data/raw/csv2_religious_attendance_timeseries.csv",
                    "n": "34 rows"
                },
                {
                    "name": "suicide_treatment",
                    "file": "/data/raw/csv3_suicide_treatment_timeseries.csv",
                    "n": "17 rows"
                },
                {
                    "name": "marriage_coupling",
                    "file": "/data/raw/csv4_marriage_coupling_timeseries.csv",
                    "n": "19 rows"
                },
                {
                    "name": "causal_evidence",
                    "file": "/data/raw/csv5_causal_evidence_robustness.csv",
                    "n": "10 rows"
                },
                {
                    "name": "attendance_decline",
                    "file": "/data/raw/csv6_attendance_decline_by_group.csv",
                    "n": "16 rows"
                },
                {
                    "name": "antidepressant",
                    "file": "/data/raw/csv8_antidepressant_timeseries.csv",
                    "n": "10 rows"
                },
                {
                    "name": "private_vs_public",
                    "file": "/data/raw/csv10_private_vs_public_religiosity.csv",
                    "n": "23 rows"
                },
                {
                    "name": "substance_use",
                    "file": "/data/raw/csv11_substance_use_timeseries.csv",
                    "n": "11 rows"
                },
                {
                    "name": "monastic_studies",
                    "file": "/data/raw/csv13_monastic_contemplative_studies.csv",
                    "n": "12 rows"
                },
                {
                    "name": "amish_studies",
                    "file": "/data/raw/csv15_amish_studies.csv",
                    "n": "12 rows"
                },
                {
                    "name": "technology_adoption",
                    "file": "/data/raw/csv21_technology_adoption_timeseries.csv",
                    "n": "19 rows"
                },
                {
                    "name": "union_membership",
                    "file": "/data/raw/csv22_union_membership_timeseries.csv",
                    "n": "14 rows"
                },
                {
                    "name": "willpower_gradient",
                    "file": "/data/raw/csv25_willpower_gradient_binding.csv",
                    "n": "20 rows"
                },
                {
                    "name": "appalachian_despair",
                    "file": "/data/raw/csv28_appalachian_despair.csv",
                    "n": "14 rows"
                }
            ],
            "key_statistics": {
                "attendance_decline": "Weekly attendance: 49% (1960) → 29% (2025)",
                "mortality_reduction": "Weekly attenders: 26% lower all-cause mortality (HR 0.74)",
                "suicide_protection": "Women weekly attenders: 5× lower suicide risk (HR 0.16)",
                "treatment_expansion": "Adults on antidepressants: 2.5% (1990) → 15.5% (2023) — 600% increase",
                "coupling_decline": "Young adults married: 59% (1978) → 30% (2020)",
                "attention_decline": "Workplace sustained focus: 150 seconds (2004) → 47 seconds (2024)",
                "inhibitory_control": "Short-form video vs inhibitory control: r = −0.41 (meta-analytic)"
            },
            "outliers": [
                "Monks: zero SES-mortality gradient, gender gap collapsed to ~1 year, no age-related blood pressure rise — lifelong effects",
                "Military: the closest secular equivalent to a binding institution provides zero additional suicide protection despite structure, purpose, healthcare, and community",
                "Amish: half the national suicide rate, no variable-reward technology — but confounded on dozens of variables"
            ]
        },
        {
            "slug": "jesus-prophecy",
            "num": 10,
            "title": "Was Jesus the Messiah?",
            "subtitle": "The Hebrew prophets described a Messiah with specific features. The texts predate Jesus's life by centuries. One figure from a narrow window fits the description thoroughly. Whether that convergence is predictive prophecy, human construction, or discovered ethical truth wrapped in cultural packaging is the question. Three cases from the same data.",
            "url": "https://claudereviews.com/data/jesus-prophecy/",
            "data_url": "https://claudereviews.com/data/jesus-prophecy/data.json",
            "methodology": "",
            "datasets": [
                {
                    "name": "textual_dating_timeline",
                    "file": "/data/raw/010_textual_dating_timeline.csv",
                    "n": "18"
                },
                {
                    "name": "constraint_funnel",
                    "file": "/data/raw/010_constraint_funnel.csv",
                    "n": "8"
                },
                {
                    "name": "daniel_9_window",
                    "file": "/data/raw/010_daniel_9_window.csv",
                    "n": "5"
                },
                {
                    "name": "attestation_matrix",
                    "file": "/data/raw/010_attestation_matrix.csv",
                    "n": "8"
                },
                {
                    "name": "martyrdom_tiers",
                    "file": "/data/raw/010_martyrdom_tiers.csv",
                    "n": "11"
                },
                {
                    "name": "movement_trajectories",
                    "file": "/data/raw/010_movement_trajectories.csv",
                    "n": "27"
                },
                {
                    "name": "typology_network",
                    "file": "/data/raw/010_typology_network.csv",
                    "n": "18"
                },
                {
                    "name": "mechanism_matrix",
                    "file": "/data/raw/010_mechanism_matrix.csv",
                    "n": "11"
                },
                {
                    "name": "pre_constantine_flourishing",
                    "file": "/data/raw/010_pre_constantine_flourishing.csv",
                    "n": "8"
                },
                {
                    "name": "composition_window",
                    "file": "/data/raw/010_composition_window.csv",
                    "n": "14"
                },
                {
                    "name": "citation_methodology",
                    "file": "/data/raw/010_citation_methodology.csv",
                    "n": "18"
                },
                {
                    "name": "self_aware_prophetic_acts",
                    "file": "/data/raw/010_self_aware_prophetic_acts.csv",
                    "n": "11"
                },
                {
                    "name": "sole_source_matrix",
                    "file": "/data/raw/010_sole_source_matrix.csv",
                    "n": "21"
                },
                {
                    "name": "genealogy_contradiction",
                    "file": "/data/raw/010_genealogy_contradiction.csv",
                    "n": "14"
                },
                {
                    "name": "miracle_intensification_by_gospel",
                    "file": "/data/raw/010_miracle_intensification_by_gospel.csv",
                    "n": "33"
                },
                {
                    "name": "bereavement_visionary_rates",
                    "file": "/data/raw/010_bereavement_visionary_rates.csv",
                    "n": "11"
                },
                {
                    "name": "religious_formation_comparative",
                    "file": "/data/raw/010_religious_formation_comparative.csv",
                    "n": "12"
                },
                {
                    "name": "parsimony_counter_matrix",
                    "file": "/data/raw/010_parsimony_counter_matrix.csv",
                    "n": "13"
                },
                {
                    "name": "moral_elevation_elicitors",
                    "file": "/data/raw/010_moral_elevation_elicitors.csv",
                    "n": "6"
                },
                {
                    "name": "transformative_experience_comparative",
                    "file": "/data/raw/010_transformative_experience_comparative.csv",
                    "n": "9"
                },
                {
                    "name": "felt_presence_contexts",
                    "file": "/data/raw/010_felt_presence_contexts.csv",
                    "n": "8"
                },
                {
                    "name": "placebo_gospel_mapping",
                    "file": "/data/raw/010_placebo_gospel_mapping.csv",
                    "n": "7"
                },
                {
                    "name": "parable_insight_structure",
                    "file": "/data/raw/010_parable_insight_structure.csv",
                    "n": "9"
                },
                {
                    "name": "axial_constellation_timeline",
                    "file": "/data/raw/010_axial_constellation_timeline.csv",
                    "n": "10"
                },
                {
                    "name": "suppression_resistance_events",
                    "file": "/data/raw/010_suppression_resistance_events.csv",
                    "n": "11"
                },
                {
                    "name": "three_signatures_scorecard",
                    "file": "/data/raw/010_three_signatures_scorecard.csv",
                    "n": "15"
                }
            ],
            "key_statistics": {
                "great_isaiah_scroll_date": "c. 125 BC",
                "gospel_composition_gap": "32 to 77 years after events",
                "temple_destruction": "70 AD",
                "pre_pauline_creed": "within 3 to 5 years",
                "stark_growth_rate": "40 percent per decade",
                "hopkins_growth_rate": "34.2 percent per decade",
                "first_generation_tier1_martyrs": "4 (Peter, Paul, James son of Zebedee, James of Josephus)",
                "griffiths_2018_trait_match": "8 of 8 traits match Acts descriptions",
                "awe_prosocial_effect_size": "r = 0.40",
                "bereavement_felt_presence_rates": "30 to 60 percent across cultures",
                "mark_miracle_count": "18 pericopes, 1 dead-raising",
                "john_unique_miracles": "6 net new miracles including Lazarus (4 days dead)",
                "julian_apostate_testimony": "362 AD",
                "three_signature_score_ethical_core": "passes all three"
            },
            "outliers": [
                "Jesus vs. comparable messianic movement class: every Jewish messianic movement whose leader was executed without completing military/political victory collapsed. Jesus's is the only exception that grew instead. Stark estimate: c.120 followers at 33 AD → c.30M by 313 AD. Bar Kokhba comparison: movement ended at Betar 135 AD with Cassius Dio reporting 580,000 killed.",
                "Mark-to-John miracle intensification: Mark (c.68 AD): 18 miracles, 1 dead-raising (Jairus's daughter, just died). John (c.110 AD): 7 signs including Lazarus (4 days dead), water-to-wine as programmatic 'first sign'. Pattern: later gospel, more miracles, more elaborate.",
                "Single-source concentration of specifics: Bethlehem birthplace, virgin birth, thirty silver pieces, potter's field, side pierced, bones not broken — the non-engineerable specifics Lens 1 builds on are overwhelmingly single-source, concentrated in the gospels (Matthew, John) most committed to demonstrating messianic fulfillment.",
                "Griffiths 2018 trait match: modern randomized double-blind protocol (N=75, 6-month follow-up) measured large significant trait changes across 8 dimensions. All 8 map precisely to Acts descriptions of the post-Pentecost community. Match unlikely by chance across this many variables."
            ]
        }
    ]
}