Reminder about potentially serious problems with a type of blocked ANOVA analysis

Steve P. Verrill; David E. Kretschmann

doi:10.2737/FPL-RP-683

Treesearch

Reminder about potentially serious problems with a type of blocked ANOVA analysis

Informally Refereed

Authors:

Steve P. Verrill, David E. Kretschmann

Year:

2017

Type:

Research Paper

Station:

Forest Products Laboratory

DOI:

https://doi.org/10.2737/FPL-RP-683

Source:

Research Paper, FPL-RP-683. Madison, WI: USDA, Forest Service, Forest Products Laboratory. 117 p.

Download (PDF 1.51 MB)

Download (PDF 1.51 MB): https://research.fs.usda.gov/download/treesearch/54781.pdf

Abstract

A type of blocked experiment has the potential of being poorly designed and/or analyzed. Verrill (1993, 1999) and Verrill et al. (2004) referred to such an experiment as a “predictor sort” experiment. David and Gunnink (1997) described the procedure as “artificial pairing.” In textbooks it is sometimes referred to as a “matched pair” or a “matched subjects” design. The associated design process is also sometimes described as “forming blocks via a comcomitant variable.” In a wood research context, Warren and Madsen (1977) described the allocation procedure as follows:

Keywords

Predictor sort sampling, artificial pairing, matched pairs, matched subjects, concomitant variable, blocked ANOVA, analysis of covariance, maximum likelihood estimation

Citation

Verrill, Steve; Kretschmann, David E. 2017. Reminder about potentially serious problems with a type of blocked ANOVA analysis. Research Paper, FPL-RP-683. Madison, WI: USDA, Forest Service, Forest Products Laboratory. 117 p.