Computer-Intensive Methods in Auditing: Bootstrap Difference and Ratio Estimation

Abstract

Computer-intensive methods are a recent development in the theory of statistics with potential applicability in audit and accounting sampling. Whereas traditional sampling approaches can require complex analytics and questionable distributional assumptions, computer-intensive methods generate sampling distributions using simple, though intensive, computer computations. Generating sampling distributions empirically provides increased versatility, ease of use, and the potential for increased efficiency. However, because computer-intensive methods are sensitive to the representativeness of sample evidence, it becomes an empirical question as to whether they outperform other approaches. This st..