{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreid4klo4svdbassraqodsmiihssrrpkyu6mpyaejnqpwoi4u4nmwne",
    "uri": "at://did:plc:wwyqal4cnqhuwyacdj7rqq3n/app.bsky.feed.post/3mf6egwyeen32"
  },
  "path": "/t/statin-related-side-effects-the-recent-lancet-publication-is-biased-toward-false-negatives/28635#post_16",
  "publishedAt": "2026-02-18T09:28:18.000Z",
  "site": "https://discourse.datamethods.org",
  "tags": [
    "Neyman"
  ],
  "textContent": "I have read many methodologically excellent responses that nonetheless rest on assumptions that ignore the scenario I raised.\n\nEven in light of what I documented in my previous comment, the impact of aspects such as vast systems of conflicts of interest and ideological commitments is not a conspiracy theory, but a hypothesis to be considered - at the methodological level - as a component of the mechanism that generates the data and conclusions we produce. I argue that any response that ignores that underlying political-economic-ideological (PEI) scenario is strongly biased (just as any response that adopts the PEI scenario as the sole explanation would be).\n\nI also argue that there is a study that adopts methods heavily skewed toward false negatives and that incorporates rituals with no foundation in the literature (not even in the work of the authors who originally developed those methods and later saw them ritualized - such as Neyman, who in 1977 clarified the need to carefully choose the target hypothesis and referred to the ‘error of the first kind’ as the contextually most relevant one, not necessarily false positives). This aspect is highly problematic and cannot be ignored in light of hypotheses that include - I emphasize ‘include’ - the PEI scenario among the possible mechanisms composing the data generator.\n\nExpressions such as “solid studies” or “solid data” can only sound like political slogans when some of the fundamental methodological aspects on which that assessment of solidity rests remain unaddressed. I reiterate that this is not a conspiratorial outcry, but a call for methodological rigor that must characterize the conduct of researchers - rigor that entails considering all contextually relevant hypotheses, including uncomfortable ones. I ask for a significant effort from everyone, especially those far more competent than myself, to maintain great caution given the troubling scenarios that unfortunately affect the scientific world. Because neutrality, transparency, honesty, and integrity are essential qualities, just as much as competence (perhaps even more so nowadays, given the increasingly vast and complex web of trust on which we all depend).",
  "title": "Statin-related side effects: the recent Lancet publication is biased toward false-negatives"
}