Posted on Jupiter.ethz.ch April 2020; last accessed November 17, 2020.
ISSN/ISBN: Not available at this time. DOI: Not available at this time.
Abstract: In the effort to react appropriately to the spread of Covid-19 it is essential to have access to data concerning the effects of the virus; the two most commonly reported statistics are persons known to have tested positive for the virus (called “Cases”) and persons who have succumbed to the virus (“Deaths”). It is important that the data are timely and accurate; in the case of the latter, how can this be checked? Here we show that datasets such as the cumulative case and death data from different countries can be analysed by comparing the reported integers to the expected distribution for such rapidly-growing statistics, which are expected to have a Benford-law type distribution. Only the leading digit is of importance in such an analysis. We show that there is an expected good agreement between theory and reported values, although there are occasional strange deviations.
Bibtex:
@misc{,
AUTHOR = {Andrew Jackson and Malcolm Sambridge},
TITLE = {Benford analysis of Covid-19 data},
YEAR = {2020},
URL = {http://jupiter.ethz.ch/~ajackson/benford.pdf},
}
Reference Type: E-Print
Subject Area(s): Medical Sciences