Multiple regression and inference in ecology and conservation biology: further comments on identifying important predictor variables

摘要: Ecologists and conservation biologists frequently use multipleregression (MR) to try identify factors influencing response variables suchas species richness or occurrence. Many used regression methods maygenerate spurious results due multicollinearity. argued that there are actually two kinds of MR modelling: (1)seeking the best predictive model; (2) isolating amounts varianceattributable each predictor variable. The former has attracted most attentionwith a plethora criteria (measures model fit penalized for modelcomplexity – number parameters) Bayes-factor-based havingbeen proposed, while latter been little considered, althoughhierarchical seem promising (e.g. hierarchical partitioning). If twoapproaches agree on which retain, then it is more likelythat meaningful (of those considered) have found. Therehas problem in that, partitioning allowed rankingof by independent explanatory power, was no(statistical) way decide retain. A solution usingrandomization data matrix coupled with ispresented, as an ecological example.