Unverified50% confidenceDecision RuleExact time
The error-corrected estimator needs to incorporate the sampling error of sensitivity and false positive rate (calibration uncertainty) into the confidence interval of the final accuracy, rather than treating judge decisions as error-free ground truth.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedCalibration error should be weighted by the consequences of decisions within the uncertainty interval, not simply averaged.79% similarUnverifiedThe margin of error only reflects the uncertainty introduced by sampling and cannot encompass non-sampling errors, such as question wording bias, non-response bias, coverage bias, and social desirability bias.76% similarUnverifiedAccuracy alone cannot support gating thresholds; the decision model must be well calibrated so that the output probability values reflect true occurrence rates.75% similarUnverifiedDenser relevance judgments reduce the probability that correct results retrieved by a model are mistakenly judged as irrelevant, thereby improving evaluation reliability.75% similarVerifiedStatistically, majority voting can effectively reduce error rates only when errors are mutually independent; when errors are highly correlated, consensus instead reinforces mistakes75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1103049API
curl https://kongchang.com/api/v1/knowledge/claims/1103049MCP
get_claim(id=1103049)