One-sided truncated sequential t-test: application to natural resource sampling

Gary W. Fowler; William G. O'Regan

Treesearch

One-sided truncated sequential t-test: application to natural resource sampling

Informally Refereed

Authors:

Gary W. Fowler, William G. O'Regan

Year:

1974

Type:

Research Paper

Station:

Pacific Southwest Research Station

Source:

Res. Paper PSW-RP-100. Berkeley, CA: Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture; 17 p

Download (PDF 844 KB)

Download (PDF 844 KB): https://research.fs.usda.gov/download/treesearch/28657.pdf

Abstract

A new procedure for constructing one-sided truncated sequential t-tests and its application to natural resource sampling are described. Monte Carlo procedures were used to develop a series of one-sided truncated sequential t-tests and the associated approximations to the operating characteristic and average sample number functions. Different truncation points and decision boundary patterns were examined. The fixed sample size t-test and Barnard's open one-sided sequential t-test were compared with the new procedure. The upper one-sided test described can easily be modified to a lower one-sided test.

Keywords

biometrics, sampling, sequential sampling, t-test

Citation

Fowler, Gary W.; O''Regan, William G. 1974. One-sided truncated sequential t-test: application to natural resource sampling. Res. Paper PSW-RP-100. Berkeley, CA: Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture; 17 p