presentation given at: 49th Biometric Conference of the German Region of theInternational Biometric Society at Wuppertal, March 2003.
ISSN/ISBN: Not available at this time. DOI: Not available at this time.
Abstract: ABSTRACT: In many sets of numbers (integer or real) it is observed that the first non-zero digit is much more often 1 than 2, more often 2 than 3, and so on. This phenomenon, called "Benford's law" or "The Significant Digit Phenomenon", has been known for more than 100 years [Newcomb 1881, Benford 1938]. Hill [Hill 1995] derived theoretically that it is based on a very universal principle. There are also some applications, for example, for detecting fraud in U.S. tax data [Nigrini 1995]. We give an elementary characterization of probability distributions following the law. It is seen that log-normally distributed data approximately obey Benford's law. Further, the question is discussed whether it is possible to detect errors or manipulation in medical data sets also
Not available at this time.
Reference Type: Conference Paper
Subject Area(s): General Interest, Medical Sciences