Environmental sampling: a brief review.

Proper application of statistical principles at the onset of an environmental study can make the difference between an effective, efficient study and wasted resources. This review distills some of the thoughts current among environmental scientists from a variety of backgrounds and organizes them according to statistical principles. Collection and review of preliminary information is important for guiding sampling design. Problems must be carefully defined before rational sample selection is possible. Many potential sources of variability exist, and as many of these sources as possible should be identified, assessed, and, ideally, minimized. Sample selection itself will depend on the precise questions that are to be addressed by the study and on what assumptions can be reasonably made about structures and patterns of contaminants in the field. Careful specification of protocols for sample procurement are needed. Choices of statistical analyses will depend on how data are collected, and intended analyses should be predetermined as part of the sampling plan. Proper attention to these principles will substantially increase the quality and efficiency of an environmental sampling program.