CONFRONTING MULTICOLLINEARITY IN ECOLOGICAL MULTIPLE REGRESSION

By M. Graham. 2003

2,529 citations recorded by the source corpus.

Subfields: Ecology, Nature and Landscape Conservation, Statistics and Probability

How researchers see this paper

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

In Ecology

In Nature and Landscape Conservation

In Statistics and Probability

Citations in the corpus

Relationships between stand growth and structural diversity in spruce-dominated forests in New Brunswick, Canada. (2009, background): “Multicollinearity is encountered when a number of correlated explanatory variables exist within a regression analysis that may cause inaccurate model parameterization, decreased statistical power, and the exclusion of significant predictor variables (Graham 2003).”
MGIDI: towards an effective multivariate selection in biological experiments (2020, background): “Consequently, the multicollinearity issue (Graham, 2003; Olivoto et al., 2017), is an obstacle faced by the SH index.”
Climatic Risk Atlas of European Butterflies (2008, background): “The threshold for variable selection was a Pearson correlation coeffi cient lower than 0.3 (Graham 2003).”
Botanical richness and endemicity patterns of Borneo derived from species distribution models (2009, background): “To avoid problems such as multi-collinearity (Graham 2003), which can result in model over-fitting (Peterson et al. 2007), we reduced the number of environmental predictors.”
North Atlantic climate variation influences survival in adult fulmars (2005, background): “Multiple regression analysis is known to be hampered by collinearity among explanatory covariates (Graham 2003).”
Multiple plant traits shape the genetic basis of herbivore community assembly (2015, methodology): “When certain pairs of traits were highly correlated with each other (0 4 < |r| < 0 8), we used the residuals from a linear regression of the two traits as a new predictor variable that was no longer correlated with the other trait (Graham 2003).”
Have jellyfish in the Irish Sea benefited from climate change and overfishing? (2011, background): “A notable benefit of such an approach is that stepwise selection of explanatory variables in multiple regression analyses is robust since the independence of the variables is certain and variable selection is unbiased (Graham, 2003).”
Factors and scales potentially important for saproxylic beetles in temperate mixed oak forest (2007, background): “The VIFvalue was normally below 2.0 and never exceeded 2.4, which are considered low values and no serious collinearity ( Graham, 2003 ).”
Spatial disaggregation of complex soil map units: A decision-tree based approach in Bavarian forest soils (2012, neutral): “When working with environmental data and in particular topographical data as covariates for statistical modeling, we always have to concern multicollinearity (Graham, 2003; Hengl and MacMillan, 2009).”
The role of habitat complexity on spider communities in native alpine grasslands of New Zealand (2013, background): “By reducing data, information from such explanatory variables can be retained while avoiding collinearity between them, which can sometimes hinder interpre- tation and analysis of ecological data (Graham, 2003).”
Empirical Analyses of Plant‐Climate Relationships for the Western United States (2006, background): “1058-5893/2006/16706-0005$15.00 overfitting, and, ultimately, predictions of a precision lower than that suggested by the regression statistics (Graham 2003).”
Shifts in plant distributions in response to climate warming in a biodiversity hotspot, the Hengduan Mountains (2018, methodology): “To avoid model over-fitting due to multicollinearity of variables (Graham, 2003), a multicollinearity test was conducted using the Pearson Correlation Coefficient (r) to examine the extent of crosscorrelation (Yang, Kushwaha, Saran, Xu, & Roy, 2013), and variables with absolute cross-correlation…”
Predictors of Juvenile Survival in Birds (2013, methodology): “…approach (e.g., capture–mark–recapture [CMR]), the standard errors of the regression coefficients can be inflated, leading to correlated errors in the regression coefficients themselves, obscuring the relative importance of the predictors (Licht 1995, Graham 2003, Gimenez et al. 2009).”