How Many of a Court's Reported Judgments Still Stand
A paper from the School of Settled Matters treats the question of how many of an appellate court's 6,412 reported judgments still stand as double sampling for stratification, with a card-index annotation whose sensitivity fades with a judgment's age as the cheap phase-one variable. It gives the cost-optimal verification rates and the cost ratio above which banding the strata by age pays (about 7.2, against the registry's 40), estimates that 0.868 of the judgments stand (standard error 0.018) from 1,000 cards and 244 verified judgments, and shows by simulation that the fading removes most of what the annotation was worth and age bands recover only part of it.
| Authors | Odalys Penhallow |
|---|---|
| School | School of Settled Matters |
| Output type | Research paper |
| Published | |
| DOI | 10.5555/slop.47tb0n |
| Pages | 7 |
| Cited by | 4 outputs |
| Version | 1.0 |
| Licence | CC BY 4.0 |
Cites
Cited by
Cite as
@misc{slop_47tb0n,
author = {Odalys Penhallow},
title = {Double Sampling with a Fading Auxiliary: How Many of a Court's Reported Judgments Still Stand},
year = {2026},
publisher = {Slop University},
doi = {10.5555/slop.47tb0n},
url = {https://slop.university/outputs/slop-paper-two-phase-sampling-to-47tb0n/},
version = {1.0},
note = {Research paper},
}