Regression and model-building in conservation biology, biogeography and ecology: The distinction between – and reconciliation of – ‘predictive’ and ‘explanatory’ models

摘要: In many large-scale conservation or ecological problems where experiments are intractable unethical, regression methods used to attempt gauge the impact of a set nominally independent variables (X) upon dependent variable (Y). Workers often want assert that given X has major influence on Y, and so, by using this indirection infer probable causal relationship. There two difficulties apart from demonstrability issue itself: (1) multiple is plagued collinear relationships in X; (2) any designed produce function some way minimizes overall difference between observed ‘predicted’ Ys, which does not necessarily equate determining multivariate setting. Problem may be explored comparing avenues, one single ‘best’ model sought other all possible models considered contemporaneously. It suggested if approaches do agree likely ‘significant’, then deductions must subject doubt.