View Complete Reference

Cáceres, JLH, García, JLP, Ortiz, CMM and Dominguez, LG (2008)

First digit distribution in some biological data sets. Possible explanations for departures from Benford's Law

Electronic J Biomed 1, pp. 27-35.

ISSN/ISBN: Not available at this time. DOI: Not available at this time.



Abstract: Aim: To explore whether the first digit law (FDL) is abided by data sets from biological origin. Materials and Methods: Data were collected from different sources, including gene data length for bacteria, pre-vaccination measles incidence data and absolute values from human MEG recordings. First digit frequencies were computed and compared to predictions from FDL. Simulations included a simple model for two-dimensional epidemics spread and a randomly set upper bound model aimed to explain the behaviour of MEG data. Results: We observed that FDL is obeyed in a case of epidemic data reported at a putative focus of spread (pre-vaccination measles incidence for Preston, England). However, peculiar departures were observed for gene length distribution in microorganisms, magneto-encephalograms (MEG), and epidemic data pooled from large geographical regions. Conclusions: Simulation studies revealed that averaging data on a scenario of propagating waves can explain some of the observed distortions from FDL. This could help to understand the behaviour of epidemics data. A randomly set upper bound model (RUBM) can likely explain the observed behaviour of MEG data. Explanation for gene length data behaviour requires further theoretical work.


Bibtex:
@article{, AUTHOR = {José Luis Hernández Cáceres, Jorge Luis Pollo García, Carlos M. Martínez Ortiz, and Luis García Domínguez}, TITLE = {First digit distribution in some biological data sets. Possible explanations for departures from Benford's Law}, YEAR = {2008}, JOURNAL = {Electronic Journal of Biomedicine}, VOLUME = {1}, PAGES = {27--35}, URL = {http://www.biomed.uninet.edu/2008/n1/caceres.html}, }


Reference Type: E-Print

Subject Area(s): Statistics