APPLICATION OF INVERSE GAUSSIAN DISTRIBUTION TO OCCUPATIONAL EXPOSUREDATA

Occupational exposure concentrations are generally assumed to vary in a log-normal manner. However, because the mean of random samples from a log-normal distribution is not lognormally distributed, it cannot, i n theory, deal with the changes in the distribution of time-weighted a verage (TWA) values when the averaging time varies. On the other hand, inverse Gaussian distribution, which is similar to a log-normal one, offers the advantage of reproducibility. This paper proposes the appli cation of inverse Gaussian distribution to occupational exposure data. First, inverse Gaussian distribution was compared with the log-normal one. Second, in order to simplify the application, the percentiles of inverse Gaussian distribution with CV=0.25 to 4 were calculated. Fina lly, this distribution was applied to occupational exposure data. (C) 1997 British Occupational Hygiene Society. Published by Elsevier Scien ce Ltd.