Common Errors in Statistics (and How to Avoid Them), 4th Edition

TO LEARN MORE

Consider using LAD regression when analyzing software data sets [Miyazaki et al., 1994] or meteorological data [Mielke et al., 1996], but heed the caveats noted by Ellis [1998].

Only iteratively reweighed general Deming regression produces statistically unbiased estimates of systematic bias and reliable confidence intervals of bias. For details of the recommended technique, see Martin [2000].

Mielke and Berry [2001, Section 5.4] provide a comparison of MRPP, Cade–Richards, and OLS regression methods. Stöckl, Dewitte, and Thierpont [1998] compare ordinary linear regression, Deming regression, standardized principal component analysis, and Passing–Bablok regression.

For more on quantile regression, download Blossom and its accompanying manual from http://www.fort.usgs.gov/products/software/blossom/.

For R code to implement any of the preceding techniques, see Good [2005, 2012].

Notes

* Y|X is read as Y given X.

¹  τ is pronounced tau.

²  65 F.3d 1002, n2 (2nd Cir. 1994).

³  Quotations are from a transcript of the decision by Chelan County Superior Court Judge John Bridges, June 6, 2005.

⁴  Garza et al v. County of Los Angeles, 918 F.2d 763 (9th Cir), cert. denied.