Wolfram Library Archive

Courseware Demos MathSource Technical Notes
All Collections Articles Books Conference Proceedings

Computerized interactive stratification in statistical audits

Karl W. Heiner
Organization: State University of New York at New Paltz, USA
Journal / Anthology

Mathematics with Vision: Proceedings of the First International Mathematica Symposium
Year: 1995
Page range: 199-206

Statistical sampling and estimation procedures are used to settle accounts between two, sometimes adversarial, parties. Two situations are described where Mathematica has been used to assist in developing optimal sampling plans. The first situation involves choosing a sample size that maximizes the financial return to the auditing party when simple random sampling is appropriate. In the discussion of the first situation, certain statistical auditing background issues are raised. We then move to a more complicated problem, where the sample size is fixed, but the book value of the population of transactions is known as well as the probability that the transaction will be determined to have been paid in error. Here, stratification may be used to substantially reduce the variance of the estimation and eventually increase the recovery. The main statistical issues will be how many strata should be used and where should the strata boundaries be drawn.

*Mathematics > Probability and Statistics