{
  "researchDate": "2026-10-07",
  "revision": "draft-3",
  "locale": "en",
  "presentation": "overview",
  "assessment": "Qualified explanation · research draft",
  "reviewNeeds": "Source verification, subject-matter and editorial approval pending.",
  "updateTriggers": "New studies, corrected sources or evidence changing the stated finding.",
  "related": [
    {
      "slug": "mmr",
      "label": "Measles vaccines: medical evidence"
    },
    {
      "slug": "misinformation-disinformation",
      "label": "Misinformation and disinformation"
    }
  ],
  "id": "vaccine-twitter-amplification-2014-2017",
  "slug": "vaccine-amplification",
  "kind": "explainer",
  "topic": "How information travels",
  "tag": "Historical vaccine debate",
  "title": "Who amplified the vaccine debate?",
  "question": "What does historical Twitter research establish?",
  "summary": "A 2018 study found Russian-troll accounts posting both for and against vaccines on Twitter in 2014–2017. It shows how some accounts helped spread the debate, not whether vaccines are safe.",
  "scope": "This study covered English-language Twitter from July 2014 to September 2017, before COVID.",
  "geographicScope": {
    "kind": "undetermined",
    "countryCodes": [],
    "jurisdictionCodes": [],
    "note": "English-language Twitter sample; not a representative national population or a current cross-platform estimate."
  },
  "evidence": [
    {
      "heading": "Accounts posted opposing messages",
      "text": "Researchers examined about 1.79 million tweets, plus 253 messages using #VaccinateUS. Accounts identified as Russian trolls posted on both sides of the vaccine debate. In the small hashtag sample, 43% of messages favored vaccination, 38% opposed it and 19% were neutral. These are percentages of those messages, not of the public.",
      "sources": [
        1
      ],
      "sourceIds": [
        "broniatowski-2018"
      ]
    },
    {
      "heading": "Posting is not the same as persuading",
      "text": "Researchers used existing lists and account classifications to identify account types. The study did not establish who ran every unknown account or whether the messages changed people’s beliefs or vaccination choices.",
      "sources": [
        1
      ],
      "sourceIds": [
        "broniatowski-2018"
      ]
    },
    {
      "heading": "A message still needs to be checked",
      "text": "An organized campaign can share a true report, a real concern or a false claim. Medical evidence still answers vaccine-safety questions. This pre-COVID study cannot tell us where COVID claims began. It does not make ordinary people who discuss vaccines foreign agents.",
      "sources": [
        1
      ],
      "sourceIds": [
        "broniatowski-2018"
      ]
    }
  ],
  "takeaway": "Check what the message says and who helped spread it as separate questions.",
  "limits": [
    "This is one historical Twitter study, not a count of all foreign influence or all vaccine discussion today.",
    "It shows observed posting, not a measured effect on vaccination or beliefs.",
    "Researchers depended on account lists they did not create themselves."
  ],
  "contextTitle": "Spreading a debate is different from starting it",
  "contextIntro": "People can exploit disagreements that already exist.",
  "detail": "A foreign-linked account sharing a viewpoint does not show that the viewpoint began abroad. Evidence about who ran an account, what they intended, whether a statement was true and whether anyone was persuaded answers different questions.",
  "sources": [
    {
      "id": "broniatowski-2018",
      "revision": "research-2026-10-07",
      "title": "Weaponized Health Communication: Twitter Bots and Russian Trolls Amplify the Vaccine Debate",
      "publisher": "American Journal of Public Health, 2018",
      "url": "https://portal.ct.gov/-/media/dcf/agency/jic-resources/ajph2018304567.pdf",
      "type": "Peer-reviewed observational content analysis",
      "access": "Seven-page primary paper accessible; selected methods/results inspected in supplied dossier, printed pp. 1381–1383",
      "limit": "July 2014–September 2017; account-list/classification limits; no measured uptake effect or COVID-origin inference.",
      "retrievedAt": "2026-10-06",
      "locale": "en",
      "geographicCoverage": {
        "kind": "undetermined",
        "countryCodes": [],
        "jurisdictionCodes": [],
        "note": "July 2014–September 2017; account-list/classification limits; no measured uptake effect or COVID-origin inference."
      },
      "readerTitle": "Twitter accounts posting on both sides of the vaccine debate",
      "readerAccess": "We could read the paper and checked its main methods and results.",
      "readerLimit": "It covers July 2014–September 2017. Researchers relied on account lists and classifications. It did not measure changed vaccination choices and cannot establish where COVID claims began.",
      "supports": [
        "Accounts posted opposing messages",
        "Posting is not the same as persuading",
        "A message still needs to be checked"
      ],
      "limitationsInternal": [
        "Seven-page primary paper accessible; selected methods/results inspected in supplied dossier, printed pp. 1381–1383",
        "July 2014–September 2017; account-list/classification limits; no measured uptake effect or COVID-origin inference."
      ],
      "limitationsPublic": [
        "It covers July 2014–September 2017. Researchers relied on account lists and classifications. It did not measure changed vaccination choices and cannot establish where COVID claims began."
      ]
    }
  ],
  "dossier": "foreign_influence_campaigns_review.md",
  "copyReviewedAt": "2026-10-07",
  "readerCopyEdition": "plain-language-1",
  "citationDesignEdition": "publication-first-1",
  "assetDesignEdition": "space-first-1",
  "headline": {
    "text": "Who amplified the vaccine debate?",
    "revision": "draft-3",
    "status": "draft",
    "sourceMessageId": null
  },
  "language": "en",
  "jurisdiction": {
    "kind": "undetermined",
    "countryCodes": [],
    "jurisdictionCodes": [],
    "note": "English-language Twitter sample; not a representative national population or a current cross-platform estimate."
  },
  "claimScope": "This study covered English-language Twitter from July 2014 to September 2017, before COVID.",
  "research": {
    "dossierRef": "docs/research/foreign_influence_campaigns_review.md",
    "reviewedAt": "2026-10-07",
    "reviewDue": "New studies, corrected sources or evidence changing the stated finding.",
    "contraryEvidenceConsidered": [],
    "status": "research draft; editorial/context review separate"
  },
  "status": "researched draft preview",
  "realWorldClaimReviewed": false,
  "clinicalEditorialApproved": false
}