|
Activity Number:
|
29
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 2, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #305131 |
|
Title:
|
Distributions That Approximately Follow Benford's and Other Digit Laws
|
|
Author(s):
|
Adam Petrie*+
|
|
Companies:
|
The University of Tennessee
|
|
Address:
|
SMC 338, Knoxville, TN, 37996,
|
|
Keywords:
|
Benford ; digit law ; lognormal ; extreme value ; Pareto ; log-logistic
|
|
Abstract:
|
Benford's Law states that, for many data sets, the probability that a number's first digit is d is equal to log10(1+1/d). Similar laws describe the distribution of any particular digit and the joint distribution of any number of digits. While some probability distributions like the lognormal almost exactly follow the first digit law, no distribution can satisfy all of them. We survey a variety of distributions for concordance with both first and second digit laws and find that many distributions commonly used in statistical modeling give digit frequencies which are statistically indistinguishable from predicted frequencies for common sample sizes. We also show that the distribution of US box office grosses is approximately Benford.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2009 program |
