Talk:Wuhan Wansheng Biotechnology

Request for comment. This organisation has by far the largest associated third-party report set on the wiki. Earlier revisions of the article described this as a strong testing record. It has also been described, in edits since reverted, as evidence of quality problems on the basis of the low tail.

Question: what, if anything, does a large self-selected report set support? Please address the selection mechanism rather than the totals. — — RfC_Rasmus (talk) 10:14, 17 June 2025 (UTC)

Almost nothing about the material, and something about the documentation. The set is self-selected in at least two opposing directions and neither bias is quantified, so the central tendency is uninterpretable. What is interpretable is dispersion relative to organisations whose certificates are standardised — and dispersion is a property of reports, which is exactly what we have. — — AnalyticalAnnie (talk) 13:42, 17 June 2025 (UTC)

Agree with Annie. I would add the base-rate point explicitly, because both sides of this dispute have got it wrong in the past: a larger set contains more low results and more high results. Anyone arguing from the count of low results alone is arguing from the size of the set. — — TallySheetTad (talk) 09:27, 18 June 2025 (UTC)

Nothing, and the article should say so at length. This is the one supplier article where the interesting content is epistemological rather than descriptive. A reader who leaves understanding why the large data set does not settle the question has learned something they can apply to the other nineteen. — — NPOV_Nadia (talk) 11:03, 19 June 2025 (UTC)

Nothing — and both previous framings should be treated as neutrality violations, not merely as errors. "Strong testing record" and "evidence of problems" are the same mistake with opposite signs. — — PuffRemoverPax (talk) 15:31, 20 June 2025 (UTC)

Nothing about quality. One caveat: the existence of a large set is itself a fact about the organisation's market position and about purchaser behaviour towards it, and that is worth a sentence. It just is not a quality finding. — — VerifyFirstVida (talk) 09:48, 23 June 2025 (UTC)

Closing. Clear consensus: the set supports no conclusion about material quality; it supports a limited comparative statement about dispersion of reported figures; and its existence is a notable fact in itself. Both prior framings are to be treated as neutrality problems. Written into §Independent testing record in those terms. — — RfC_Rasmus (talk) 10:52, 28 June 2025 (UTC)

Placing the standing note at the top of every supplier talk page so it does not have to be re-argued. Under Project:Vendor notability a supplier article may not contain: a price, an offer, a discount, a price list, any contact detail, or a shipping-outcome claim not attributed to a source.

Additions of any of those are removed on sight and are not a content dispute. — — NPOV_Nadia (talk) 09:12, 4 July 2024 (UTC)

Support, and worth adding the reason rather than only the rule: each of those items turns a reference entry into a sales listing, and the moment one article carries them every other organisation has a grievance. — — CoI_Corin (talk) 11:48, 4 July 2024 (UTC)

Support. I would add "adjectives not attributable to a source" to the list, which is the softer version of the same problem and much the commonest. — — Hedgerow_Hal (talk) 10:23, 5 July 2024 (UTC)

Added to the standing note, so that it does not have to be argued again on each of the twenty pages. ✓ Resolved — — NPOV_Nadia (talk) 08:57, 6 July 2024 (UTC)