Statistics for Censored Environmental Data by Dennis R. Helsel
The measurement of trace chemicals in water, air, soil, and biota frequently results in values reported only as less than the detection limit.
The standard practice in environmental science for interpreting these data is woefully inadequate, usually consisting of substituting one-half the detection limit and performing traditional statistics tests.
Offering a unique approach, Nondetects and Data Analysis presents methods that are widely used in medical and industrial settings but have been rarely or never applied to issues of environmental health.
The steps necessary to take these existing procedures and apply them to left-censored environmental data are fully illustrated.
Contents:
![]()
|