Abstract
The natural complexity of ecological communities regularly lures ecologists to collect elaborate data sets in which confounding factors are often present. Although multiple regression is commonly used in such cases to test the individual effects of many explanatory variables on a continuous response, the inherent collinearity (multicollinearity) of confounded explanatory variables encumbers analyses and threatens their statistical and inferential interpretation. Using numerical simulations, I quantified the impact of multicollinearity on ecological multiple regression and found that even low levels of collinearity bias analyses (r ≥ 0.28 or r2 ≥ 0.08), causing (1) inaccurate model parameterization, (2) decreased statistical power, and (3) exclusion of significant predictor variables during model creation. Then, using real ecological data, I demonstrated the utility of various statistical techniques for enhancing the reliability and interpretation of ecological multiple regression in the presence of multicollinearity.
Keywords
MulticollinearityCollinearityStatisticsRegression analysisVariance inflation factorEcologyConfoundingLinear regressionEconometricsRegressionRegression diagnosticMathematicsBiologyPolynomial regression
Affiliated Institutions
Related Publications
Multicollinearity
Abstract Multicollinearity refers to the linear relation among two or more variables. It is a data problem which may cause serious difficulty with the reliability of the estimat...
Collinearity: a review of methods to deal with it and a simulation study evaluating their performance
Collinearity refers to the non independence of predictor variables, usually in a regression‐type analysis. It is a common feature of any descriptive ecological data set and can ...
Multimodel inference in ecology and evolution: challenges and solutions
Information theoretic approaches and model averaging are increasing in popularity, but this approach can be difficult to apply to the realistic, complex models that typify many ...
On the misuse of residuals in ecology: regression of residuals vs. multiple regression
1 Residuals from linear regressions are used frequently in statistical analysis, often for the purpose of controlling for unwanted effects in multivariable datasets. This paper ...
A Comparison of Least Squares and Latent Root Regression Estimators
Miilticollinesrity among the columns of regressor variables is known to cause severe distortion of the least squares estimates of the parameters in a multiple linear regression ...
Publication Info
- Year
- 2003
- Type
- article
- Volume
- 84
- Issue
- 11
- Pages
- 2809-2815
- Citations
- 2447
- Access
- Closed
External Links
Social Impact
Altmetric
PlumX Metrics
Social media, news, blog, policy document mentions
Citation Metrics
2447
OpenAlex
Cite This
Michael H. Graham
(2003).
CONFRONTING MULTICOLLINEARITY IN ECOLOGICAL MULTIPLE REGRESSION.
Ecology
, 84
(11)
, 2809-2815.
https://doi.org/10.1890/02-3114
Identifiers
- DOI
- 10.1890/02-3114