Article

Journal of Agricultural, Biological, and Environmental Statistics

, Volume 7, Issue 3, pp 300-305

Statistical methods helping and hindering environmental science and management

  • Graham B. McBrideAffiliated withNIWA (National Institute of Water and Atmospheric Research) Email author 

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Environmental scientists face the reality that many of their journals’ editors and referees routinely insist that results be accompanied by statements of statistical significance, obtained from two-sided tests of point-null hypotheses. Many in these three groups of people appear only vaguely a ware of the arbitrarinessoften invoked by this procedure and of the information sterility in a single p-value. The interpretation to be made of the failure of a test to attain such significance is not clear. For such reasons, some colleagues (and senior statisticans) have called current usage of the procedures into serious question. Some reasons for this dislocation and some of the more dramatic consequences for environmental science and management are presented. Interval and Bayesian approaches can offer remedies.

Key Words

Burden of proof Compliance rules Credibility and confidence intervals Equivalence Interval hypothesis Null hypothesis