Statistical framework for comparing lumber sorting procedures

Steve P. Verrill; Richard A. Johnson; James W. Evans

doi:10.2737/FPL-RP-651

Treesearch

Statistical framework for comparing lumber sorting procedures

Informally Refereed

Authors:

Steve P. Verrill, Richard A. Johnson, James W. Evans

Year:

2009

Type:

Research Paper

Station:

Forest Products Laboratory

DOI:

https://doi.org/10.2737/FPL-RP-651

Source:

Research Paper FPL-RP-651

Download (PDF 507 KB)

Download (PDF 507 KB): https://research.fs.usda.gov/download/treesearch/33886.pdf

Abstract

In this paper, we take a preliminary look at techniques for comparing sorts that yield grades of lumber. We propose methodology that takes into account differences in grade prices and the costs associated with misgrading. We focus on two- and three-category sorts, but our results could be readily extended. We provide web links to sample FORTRAN implementations of this methodology for the case in which the strength predictor and strength have a bivariate normal distribution and the load distribution is also normally distributed. We indicate how these approaches would have to be modified for other predictor, strength, and load distributions.

Keywords

Lumber, grading, mathematical models, computer simulation, lumber grading, correlation estimation, comparing sorting procedures, classification

Citation

Verrill, Steve V.; Johnson, Richard A.; Evans, James W. 2009. Statistical framework for comparing lumber sorting procedures. Research paper FPL-RP-651. Madison, WI: U.S.Department of Agriculture, Forest Service, Forest Products Laboratory. 17 p.