Large studies find no link between vaccines and autism.

1. Pass Title matches the evidence

   Title that large studies find no link between vaccines and autism matches the cohort, meta-analysis and WHO evidence and the False finding.

2. Pass No obvious contradictions

   Finding False, answer, and conclusion agree; the confidence and scope limits are stated consistently.

3. Pass Plausible justification

   Conclusion explains the False rating from converging large cohorts, a meta-analysis and an international review, with population-level limits stated.

4. Pass Sources are relevant

   Hviid et al., Jain et al., Taylor et al. and the WHO GACVS statement are on topic for vaccines and autism by name and description.

5. Pass Common questions answered

   Three common questions are present and each response plausibly addresses its question.

Coverage: Read the full article payload; sources judged by name/title/description/role only, per First Look's rules; no source snapshots read.

1. Concern (minor) Central claims are accurate

   Correct evidence e4: WHO's 31 studies were not all from 11 countries; 20 of the 31 (from 11 countries) found no association, while 11 low-quality studies suggested one.

   Summary: the article rates 'Vaccines cause autism' False (High), resting on two MMR cohorts, a meta-analysis and the 2025 WHO GACVS statement. Claim ratings: (1) Finding False / answer 'no link' - True; WHO source: available high-quality evidence 'indicates that vaccines... do not cause autism'. (2) Danish cohort 657,461 children, HR 0.93, CI including 1 - Unverifiable from source (403), consistent with the article's DOI and auditor knowledge. (3) Jain >95,000 US children incl. older-sibling-with-autism group, no harmful association - Unverifiable (403), consistent with knowledge. (4) Taylor 2014 meta-analysis found no association incl. MMR/thimerosal - Unverifiable (empty text), consistent with title. (5) WHO 'assessment of 31 studies from 11 countries' - Mostly True: the source says 31 primary studies were included, but 'Twenty of the 31 primary research papers originating from 11 different countries' found no association; 'The other eleven studies (with nine originating from one single research group in the United States) suggested a potential association' with high risk of bias. The article conflates the 31 and the 11 countries and implies all 31 support no link. Peripheral, does not change the core message, so Minor. Confidence: Medium (three sources unreadable).

   About evidence:international-safety-review, source:who2025 The finding

2. Concern (minor) Context is complete

   Mention that WHO judged the 11 of 31 studies suggesting a link to be at high risk of bias, and cite its aluminium-adjuvant review, since both cohorts cover only MMR.

   The cited WHO statement explicitly weighs contrary evidence: eleven of 31 studies 'suggested a potential association; however, these had multiple and significant methodological issues... high risk of bias', and two ecological aluminium studies reported an association but were 'judged at critical risk of bias'. The article never acknowledges that any studies report an association, which is the evidence proponents cite. It also omits WHO's second review on aluminium-adjuvanted vaccines (10 RCTs, 7 cohorts, no association) even though its two cohort items (e1, e2) concern only MMR while the claim covers all vaccines. Because WHO discounts the contrary studies and the meta-analysis covers thimerosal, a reader's overall impression would not change, so Minor. Confidence: High.

   About evidence:international-safety-review, source:who2025, conclusion The finding

3. Pass Claims are supported and labeled

   Every key claim is tied to an evidence item with a cited source (hviid2019, jain2015, taylor2014, who2025). The conclusion separates the factual finding from its inference: 'concerns population evidence about causal association; it is not an explanation or diagnosis for an individual'. No opinion is presented as fact. Confidence: High.

4. Pass Sourcing is strong and transparent

   Sources are two peer-reviewed cohort studies with DOIs (Annals of Internal Medicine, JAMA), a peer-reviewed meta-analysis (Vaccine) and a primary WHO committee statement. Each evidence item names its source id, so it is clear which source supports which claim. No partisan, anonymous or circular sourcing. Note: the three DOI sources have no archived copy and two block automated access, which limits reader verification but is not a quality problem. Confidence: High.

5. Pass Framing is fair and neutral

   The claim is stated as it circulates ('Vaccines cause autism.'). Objections are stated fairly and answered with evidence, e.g. 'Timing alone cannot distinguish coincidence from causation', and the article concedes 'these designs cannot establish impossibility'. No ridicule or loaded language. The image caption is neutral. Confidence: High.

6. Pass Uncertainty is stated

   Each evidence item lists specific limitations (registry coding and residual confounding; claims-data misclassification and subgroup imprecision; meta-analysis inherits shared weaknesses; committee assessment depends on included studies). The conclusion states 'does not claim that science can prove absolute impossibility'. High confidence is justified by converging large cohorts and two syntheses. The omitted conflicting low-quality studies are reported under context-and-omissions rather than counted twice. Confidence: High.

Coverage: Read the full article payload. Source snapshots: who2025 read in full (live copy). hviid2019 and jain2015 could not be read (HTTP 403 from publisher, no archive or Wayback capture). taylor2014 resolved (HTTP 200) but returned empty text. Claims resting on those three sources were checked only against the article's own descriptions and the auditor's background knowledge (consistent, but not verified against the source text); no Fail rests on them. Recommend archiving the three DOI sources.

1. Concern (material) Strongest objections answered

   Add an objection addressing whether non-MMR exposures (aluminium adjuvants, thimerosal, the cumulative early schedule) cause autism, with evidence beyond the two MMR cohorts.

   The objections answered are timing ('Autism can be noticed after a vaccination'), family risk, and 'cannot prove that risk is exactly zero'. These are answered fairly with evidence. However, the claim is about vaccines generally, and both cohort items (e1, e2) concern only MMR. The article does not address the version of the claim that currently circulates most strongly: that aluminium adjuvants, thimerosal, or the cumulative number of early-childhood vaccines (rather than MMR) cause autism. Only the meta-analysis summary mentions thimerosal in passing, and nothing addresses adjuvants or schedule. Because this is the main form of the claim proponents now advance, its absence from the objections is the strongest challenge left unanswered; it would change how strongly a reader can apply the conclusion to 'vaccines' as a class. Rests on auditor knowledge, so Material at most.

   About conclusion, evidence:danish-registry-cohort, evidence:us-family-risk-cohort The finding

2. Concern (minor) Evidence is independent

   Note that the WHO 2025 review (2010-2025 studies) and Taylor 2014 likely include or overlap the Danish and US cohorts, so the four items are not fully independent.

   Method step 2 presents 'two cohorts, a meta-analysis, and an international assessment' as evidence that 'does not depend on one institution or method'. The two cohorts are genuinely independent (different countries, data systems, institutions). But the WHO assessment (e4) synthesizes '31 studies... together with meta-analyses', and the article does not address whether these include Jain 2015, Hviid 2019, and the Taylor 2014 meta-analysis itself; given the years, they very likely do. Taylor 2014 likewise pools earlier cohorts, possibly including earlier Danish registry data. So e3 and e4 partly re-count e1 and e2. Independent support from the two cohorts remains, so Minor.

   About evidence:international-safety-review, evidence:pooled-vaccine-analysis, evidence:danish-registry-cohort, evidence:us-family-risk-cohort The finding

3. Pass Rates the real claim

   Claim 'Vaccines cause autism.' is the claim as it circulates, not a strawman. Method step 1 defines it as 'a causal association between vaccination and autism', which matches what a reader repeating it means. The conclusion stays within scope: population evidence on causal association, explicitly not individual diagnosis. Scope across vaccine types is addressed under strongest-objection rather than here. I would rate the same claim.

4. Pass Rating follows from the evidence

   Before reading the verdict I would rate False: two large cohorts (657,461 Danish children, HR 0.93 with CI including 1; >95,000 US children incl. higher-family-risk group), a meta-analysis reporting 'no association between vaccination, including MMR or thimerosal exposure, and autism', and a WHO committee concluding evidence 'does not support a causal relationship'. The article's finding False and answer 'found no link' match. No contradiction between answer and finding.

5. Pass Confidence matches the evidence

   High confidence rests on multiple large, independent-method studies across countries plus two syntheses; I would also give High. Title 'Large studies find no link between vaccines and autism.' and summary do not overstate: the conclusion explicitly says it 'does not claim that science can prove absolute impossibility'. Limitations that matter (observational design, residual confounding, claims-data misclassification, shared weaknesses in pooling) are stated per item.

Coverage: Read the full article payload (claim, finding, answer, confidence, summary, context, conclusion, four evidence items with limitations, three objections, five method steps, four sources). Per the Skeptic's instructions, no source snapshots were read; evidence was taken as stated.

1. Concern (minor) Title states the finding

   Rewrite seoTitle to state the finding, e.g. "Do Vaccines Cause Autism? Large Studies Find No Link".

   Title 'Large studies find no link between vaccines and autism.' names the subject, resolves the claim, and does not repeat the myth; its negative form describes an absence of evidence, which the rule permits. Rule: 'seoTitle may use the wording people search with as long as it states the finding.' seoTitle 'Do Vaccines Cause Autism? What the Evidence Shows' poses the question but does not state the finding. Minor per the rule ('a seoTitle that does not state the finding').

   About conclusion The finding

2. Concern (minor) Verdict fields agree and stand alone

   Replace "is not supported by" in summary and "does not support" in conclusion with wording matching the False label, e.g. "The claim is false: converging evidence...".

   Verdict read: FactHub rates the claim "Vaccines cause autism." False (High confidence): "Large cohort studies, including 657,461 Danish children, found no link between vaccination and autism." The claim is worded as it circulates, and the answer stands alone. However, the changelog records the label changed from "Not supported" to False, and the summary still opens 'The claim is not supported by converging evidence' and the conclusion 'The best available evidence does not support the claim'. Rule: all fields must express the conclusion 'in the current label's terms'; stale wording from an earlier label is Minor.

   About conclusion The finding

3. Concern (minor) Each section does its job

   Cut summary to one sentence, and drop the hand-written independence analysis from context and method step 5; the page generates it.

   Rules: summary is one sentence naming the strongest independent lines of evidence; context is essential framing only; the author must not write an independence note. The summary is two sentences. Context ends 'The conclusion is strong because evidence from different methods, institutions, countries, and study periods converges' and method step 5 reads 'Counted four distinct methods, evidence from multiple countries, and studies or assessments spanning the 2010s and 2020s' - both hand-written independence notes repeating what the page generates. The conclusion does synthesize (it explains convergence, scope, and what remains unproven), and editorialNote states scope, caveats, and review status ('No external specialist review has been recorded'), so these are Minor misplacements.

   About conclusion, method:5 The finding

4. Concern (minor) Limitations are specific

   Make the meta-analysis and WHO limitations item-specific, e.g. say which vaccines and study designs Taylor 2014 pooled and what that leaves unproven.

   Rule: limitations are specific to the item. The two cohort limitations are specific (registry coding and residual confounding; managed-care claims data, misclassification, subgroup imprecision). The meta-analysis limitation ('depends on the design, reporting, comparability, and bias of its included studies') and the WHO limitation ('depends on the completeness and quality of the included studies') would apply to any synthesis and do not say what these particular items leave unproven. These are supporting syntheses alongside the two cohorts, so Minor.

   About limitation:pooled-vaccine-analysis, limitation:international-safety-review The finding

5. Concern (minor) Consistent house style

   Spell out MMR (measles, mumps, and rubella) at first use and briefly explain "hazard ratio" and "confidence interval" for general readers.

   Rules: 'an abbreviation is spelled out at first use' and 'technical terms a general reader would not know are explained at first use'. 'MMR' appears in keyPoints and evidence titles/summaries without being spelled out in reader text (only the source title expands it). 'adjusted hazard ratio' and 'confidence interval' are used without explanation. Numbers (657,461; 95,000), names, and American spelling are otherwise consistent; no drafting text found. Minor.

   About evidence:danish-registry-cohort, conclusion The finding

6. Pass Tone is measured and respectful

   Language is measured: 'does not support', 'does not claim that science can prove absolute impossibility'. Objections are answered respectfully and with evidence, e.g. 'That is correct: these designs cannot establish impossibility.' No ridicule, sarcasm, or unwarranted absolutes.

7. Pass Readers can verify it

   Sources have role labels ('Nationwide cohort study', 'Family-risk cohort study', 'Meta-analysis', 'International scientific assessment'), full titles, DOIs for the three studies and a stable WHO URL, plus short descriptions. Method steps cover claim definition (step 1), evidence not resting on one source (step 2), sources reviewed with identifiers (step 3), and recording of limits (step 4). Key claims have primary anchors.

Coverage: Used only the article payload: title, seoTitle, description, claim, finding, answer, confidence, summary, keyPoints, context, conclusion, four evidence items and limitations, three objections, five method steps, editorialNote, four sources, image caption/alt text, and changelog. No sources or live web consulted.