Authors
S. Stanley Young, PhD, Assistant Director of Bioinformatics, National Institute of Statistical Sciences young@niss.org
Paul Fogel, PhD, Consultant, National Institute of Statistical Sciences paulfogel
The current environmental epidemiology paradigm is that air pollution is an acute killer. The head of the EPA said to a congressional committee, "Air pollution can kill you right now." Unfortunately, the EPA is relying on papers by people who most often will not provide the data sets they use to support their claims – "trust me" science. Daily data for PM2.5, ozone and deaths were obtained for the years 2007-2010 for California, three time series. There is a need for a simple and transparent way to look at this data. Two JMP scripts were written to look at this data: a moving median for a time series and a new type of p-value plot, -log10 of p-value versus expected values of order statistics from the uniform distribution. Deviations from the three time series, PM2.5, ozone, and daily deaths, were visualized and analyzed to see if increases in air pollution are associated with increases in death. The fire season in California provides "natural experiments" where air pollution increases greatly. Bottom line: there is no apparent association of air pollution with daily deaths in California.
