The bootstrap: What the government auditor should know

Many government auditors routinely use random samples to estimate some unknown amount from a population of interest. This article is intended to introduce government auditors to an alternate method of computing the lower limit that is more appropriate, provided two ordinary conditions are met. First...

Full description

Saved in:
Bibliographic Details
Published inThe journal of government financial management Vol. 51; no. 3; p. 24
Main Authors Kvanli, Alan H, Schauer, Robert
Format Journal Article
LanguageEnglish
Published Alexandria Association of Government Accountants 01.10.2002
Subjects
Auditing profession
Auditors
Audits
Confidence intervals
Estimates
Governmental accounting
Sample size
Sampling techniques
Statistical analysis
Statistical methods
United States > US
Online AccessGet full text

Cover

More Information
Summary:Many government auditors routinely use random samples to estimate some unknown amount from a population of interest. This article is intended to introduce government auditors to an alternate method of computing the lower limit that is more appropriate, provided two ordinary conditions are met. First, the samples contain only a few occurrences of the item of interest. And second, the number of items sampled as a percentage of the population is low, say less than 10%. This alternate method is called the bootstrap. The bootstrap procedure will provide more realistic or reliable - that is, not overly conservative - lower limits. Some real-life examples and computer simulations that illustrate how the bootstrap provides a higher lower limit when compared to traditional statistical computations are presented.
ISSN:1533-1385