On the misuse of residuals in ecology: regression of residuals vs. multiple regression

By R. Freckleton. 2002

470 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Health, Toxicology and Mutagenesis, Nature and Landscape Conservation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Health, Toxicology and Mutagenesis

In Nature and Landscape Conservation

Citations in the corpus

Invariant Representations without Adversarial Training (2018, methodology): “Simple statistical techniques usually involve modifying analyses with corresponding covariate effects [18] or sometimes multi-level regressions [7].”
Rural-Urban Differences in Escape Behavior of European Birds across a Latitudinal Gradient (2017, background): “Multiple regression is the best approach to control for undesirable confounding effects among correlated covariates (such as body and brain masses), yielding unbiased coefficient estimates (Freckleton, 2002).”
Fecundity Selection and the Evolution of Reproductive Output and Sex-Specific Body Size in the Liolaemus Lizard Adaptive Radiation (2011, methodology): “However, it has been shown that a number of relevant assumptions may not hold for residuals, and that the residual index is an ad-hoc sequential procedure with no demonstrated statistical justification (Garcia-Berthou 2001; Green 2001; Freckleton 2009).”
Accounting for the causal basis of collinearity when measuring the effects of habitat loss versus habitat fragmentation (2016, methodology): “Simulations have been used to demonstrate this point 409 previously (Freckleton 2002). Because of this, we suggest that researchers should think carefully 410 before following Smith et al.'s (2009) suggestion that predictors should not be removed from 411 models simply because they are collinear.”
Evading the Adversary in Invariant Representation (2018, methodology): “Simple statistical techniques usually involve modifying analyses with corresponding covariate effects [17] or sometimes multi-level regressions [7].”
Global analysis of bird elevational diversity (2009, methodology): “These corrections are not reanalyses of the area residuals from SAR regressions, so avoid pitfalls inherent in analyses of residuals (e.g. Freckleton, 2002).”
Eight (and a half) deadly sins of spatial analysis (2012, methodology): “…to remind workers of some of the pitfalls associated with multiple regression in an ecological context and to propose methods to deal with them (Freckleton, 2002, 2009; Graham, 2003; Whittingham et al., 2006), it remains that while it is a reasonable method for predicting relationships, it is…”
The paradigm of body condition: a critical reappraisal of current methods based on mass and length (2010, background): “Hence, the continuing use of ratios in fisheries and health sciences or the widespread use of Ri in terrestrial ecology, despite criticisms of these approaches (Albrecht, Gelvin & Hartman 1993; Packard & Boardman 1999; Garcı́a-Berthou 2001; Green 2001; Freckleton 2002).”
Collinearity: a review of methods to deal with it and a simulation study evaluating their performance (2013, result): “Note that while this approach is sometimes also called “residual regression” (Graham 2003), it is fundamentally different from the rightly criticised approach of 325 “regression of residuals” (Freckleton 2002).”
Sperm competition and the evolution of sperm midpiece volume in mammals (2005, methodology): “Additionally, re-examination of overall sperm length confirms the results of Gage & Freckleton’s (2003) study of mammalian sperm morphometry in which they found no correlations between linear measures of sperm and mating systems or residual testis sizes.”
Sexually size dimorphic brains and song complexity in passerine birds (2005, unknown): “An anonymous referee suggested avoiding the use of residuals, because they make an implicit assumption that independent variables are orthogonal (Freckleton, 2002).”
Size-Correction and Principal Components for Interspecific Comparative Studies (2009, methodology): “…(2001) and Freckleton (2002, 2009) both caution against using sizecorrected residuals as data for subsequent statistical analyses, and instead recommend that size be included as a covariate (Garcı́aBerthou 2001) or as an additional independent variable in a multiple regression (Freckleton 2002).”
Test of a Body Condition Index with Amphibians (2012, background): “There is a substantial record in the literature addressing the potential shortcomings of many BCIs, including various ratios, relative mass measures, and residuals indices (Jakob et al., 1996; Kotiaho, 1999; Green, 2001; Garcia-Berthou, 2001; Freckleton, 2002; Peig and Green, 2010).”
The functional response in three species of herbivorous Anatidae: effects of sward height, body mass and bill size (2003, unknown): “Although bill measurements and body size were sometimes correlated, we always included one bill measurement and body mass in the multiple regression: despite the collinearity of the two independent variables, the regression estimates are more reliable (Freckleton 2002).”