SYMPOSIA PAPER Published: 01 January 1995

Survey Design, Statistical Analysis, and Basis for Statistical Inferences in Coastal Habitat Injury Assessment: Exxon Valdez Oil Spill


The objective of the Coastal Habitat Injury Assessment study was to document and quantify injury to biota of the shallow subtidal, intertidal, and supratidal zones throughout the shoreline affected by oil or cleanup activity associated with the Exxon Valdez oil spill. The results of these studies were to be used to support the Trustee's Type B Natural Resource Damage Assessment under the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA). A probability based stratified random sample of shoreline segments was selected with probability proportional to size from each of 15 strata (5 habitat types crossed with 3 levels of potential oil impact) based on those data available in July, 1989. Three study regions were used: Prince William Sound, Cook Inlet/Kenai Peninsula, and Kodiak/Alaska Peninsula. A Geographic Information System was utilized to combine oiling and habitat data and to select the probability sample of study sites. Quasi-experiments were conducted where randomly selected oiled sites were compared to matched reference sites. Two levels of statistical inferences, philosophical bases, and limitations are discussed and illustrated with example data from the resulting studies.

Author Information

McDonald, LL
WEST, Inc., Cheyenne, WY
Erickson, WP
WEST, Inc., Cheyenne, WY
Strickland, MD
WEST, Inc., Cheyenne, WY
Price: $25.00
Contact Sales
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Developed by Committee: E47
Pages: 296–311
DOI: 10.1520/STP19868S
ISBN-EB: 978-0-8031-5293-9
ISBN-13: 978-0-8031-1896-6