STATISTICAL CONSIDERATIONS IN THE DESIGN AND ANALYSIS OF ENVIRONMENTAL-DAMAGE ASSESSMENT STUDIES

An environmental assessment of chemical and oil-spill effects presents unique statistical problems for both design and analysis. Because of the unplanned and accidental nature of a chemical spill, baseline data are usually absent and the quintessential elements of replication and randomization found in true experiments are absent. Alternative appro aches to the design of damage assessment studies are presented that in vestigate spatial and/or temporal trends in biological responses cente red about the spill event. Inferential and logistical limitations of a lternative designs are discussed along with consideration of the spati al scale for pseudo-replication and the power of tests of impact. Meta -analysis is suggested to improve both the inferential capabilities an d statistical power of damage assessments.