Journal
ACTA OECOLOGICA-INTERNATIONAL JOURNAL OF ECOLOGY
Volume 32, Issue 2, Pages 203-206Publisher
ELSEVIER
DOI: 10.1016/j.actao.2007.04.003
Keywords
hypothesis testing; power tests; confidence intervals; biological relevance; statistical significance; negative results; effect size; ecological inference
Categories
Ask authors/readers for more resources
one of the main questions that ecologists pose in their investigations includes the analysis of differences in some trait between two or more populations. I argue here that asking whether there are differences or not between populations is biologically irrelevant, since no two livings things are ever equal. On the contrary the appropriate question to pose is how large differences are between populations. That is, we urge a shift in interest from statistical significance to biological relevance for proper knowledge accumulation. I emphasise that to test biologically informative hypotheses from a classical perspective, there are two tools available: (a) the use of a priori power tests; and (b) the use of confidence intervals. Using both ensures that statistical significance and biological relevance are not decoupled, and studies yielding negative results do not need to be discarded balancing the current bias in publishing mostly 'positive' results. Complex ecological questions however require the formulation of multiple hypotheses and hence the use of modern alternative tools for ecological statistical inference such as information theoretical criteria and Bayesian statistics. (C) 2007 Elsevier Masson SAS. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available