Poisson sampling - The adjusted and unadjusted estimator revisited

Michael S. Williams; Hans T. Schreuder; Gerardo H. Terrazas

doi:10.2737/RMRS-RN-4

Treesearch

Poisson sampling - The adjusted and unadjusted estimator revisited

Informally Refereed

Authors:

Michael S. Williams, Hans T. Schreuder, Gerardo H. Terrazas

Year:

1998

Type:

Research Note

Station:

Rocky Mountain Research Station

DOI:

https://doi.org/10.2737/RMRS-RN-4

Source:

Res. Note. RMRS-RN-4. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 10 p.

Download (PDF 336 KB)

Download (PDF 336 KB): https://research.fs.usda.gov/download/treesearch/30633.pdf

Abstract

The prevailing assumption, that for Poisson sampling the adjusted estimator "Y-hat a" is always substantially more efficient than the unadjusted estimator "Y-hat u" , is shown to be incorrect. Some well known theoretical results are applicable since "Y-hat a" is a ratio-of-means estimator and "Y-hat u" a simple unbiased estimator. We formalize an additional realistic situation for high-value timber estimation for which "Y-hat u" is more efficient. (Please note: equations are spelled out inside quotation marks. Please see PDF for symbols.)

Keywords

Poisson sampling, adjusted estimator, unadjusted estimator, generalized regression estimator, approximate Srivastava estimator

Citation

Williams, Michael S.; Schreuder, Hans T.; Terrazas, Gerardo H. 1998. Poisson sampling - The adjusted and unadjusted estimator revisited. Res. Note. RMRS-RN-4. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 10 p.