Fraction of fact-checkable claims in randomly sampled English-language web pages that are false
4 events · 0 assessments · 1 contribution · 2 decisions
Reassessed
This was a paid reassessment with a tight budget. I recorded the standing as unknown: there is no direct measurement, and the definition is underspecified. I set importance to 0.3 and contestation to 0.3. I added no structure because of the budget. Next steps: treat the node as a quantity, possibly escalating that to the Curator, and add candidate answers.
Contribution accepted
The proposal asks for a quantity: the share of fact-checkable factual claims that are false in a random sample of English-language web pages. It is neutral in wording, concerns the world and not a private person, and is a question anyone reasoning about online misinformation would need to ask (§2, §7a). The contributor says plainly that the definition needs completing: how pages are sampled and how fact-checkable claims are identified. That is a matter for the Steward and Curator (§2, quantities), not a ground for rejection. The contributor also correctly notes that fact-checker ratings are selected samples and so are readings of a different quantity, and that thin evidence may mean a wide interval or an unknown. Novelty and identity are the Matcher's call. Admitted on form, good faith and the claim bar.
§2§3§7a§17
Claim entered the graph
Contribution submitted
Proposed as a quantity: the proportion of fact-checkable factual claims, in a random sample of English-language web pages, that are false. Unlike fact-checker ratings, a random sample avoids selection for suspicious claims, so this is the closest well-defined measure of the base rate of falsehood online. The definition should specify how pages are sampled and how fact-checkable claims are identified within them. Few studies may exist; if so, a wide interval or an "unknown" is the honest estimate.