{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreic3szgseqsl5cratsj7x7sgerelfqiiu2hartsffchqiwkuzqifyi",
"uri": "at://did:plc:wwyqal4cnqhuwyacdj7rqq3n/app.bsky.feed.post/3mfk5uspqgem2"
},
"path": "/t/change-the-range-not-the-language-on-confidence-intervals/27740#post_20",
"publishedAt": "2026-02-23T14:27:12.000Z",
"site": "https://discourse.datamethods.org",
"tags": [
"PDF"
],
"textContent": "What do you think of broadening the discussion from a single interval to the set of “confidence” intervals, leading to confidence distributions?\n\nI agree that the Bayesian perspective is more natural and relevant, but frequentist reports are common, and I think it is important to understand how to interpret them on their own terms.\n\nConfidence distributions, accoring to Pawitan, have an interpretation as an “extended likelihood” ie. a likelihood that partially accounts for its uncertainty. A more precise definition is given in the paper.\n\n**Pawitan, Y., & Lee, Y. (2021).** Confidence as likelihood. Statistical Science, 36(4), 509-517. (PDF)",
"title": "Change the range not the language on confidence intervals"
}