Inference on Rare Errors Using Asymptotic Expansions and Bootstrap Calibration.211 Probability, Networks and Algorithms

摘要

The number of items in error in an audit population is usually quite small,211u001ewhereas the error distribution is typically highly skewed to the right. For 211u001eapplications in statistical auditing, where line item sampling is appropriate, a 211u001enew upper confidence limit for the total error amount in an audit population is 211u001eobtained. Our method involves an empirical Cornish-Fisher expansion in the first 211u001eplace; in the second stage we employ the bootstrap to calibrate the coverage 211u001eprobability of the resulting interval estimate.