Abstract
The Forest Inventory and Analysis (FIA) Program uses stratified estimation techniques to produce estimates of forest attributes. Stratification must be unbiased and stratification procedures should be examined to identify any potential bias. This note explains simple techniques for identifying potential bias, discriminating between sample bias and stratification bias, and determining the magnitude of the effect of stratification bias on forest area estimates. The key recommendation is that checks and balances should be incorporated into the FIA processing system to reduce the likelihood of bias caused by stratification.
Keywords
bias,
FIA,
national land cover dataset,
photointerpretation,
proportional allocation
Citation
Coulston, John. 2008. Forest inventory and stratified estimation: a cautionary note. Res. Note SRS-16. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southeastern Forest Experiment Station. 8 p.