Two new papers on statistical methods are now available on “Early View”.
The first is a paper by Holger Schielzeth on regression methods. In this paper Holger tries to raise awareness of the importance of standardizing and centring input variables before analysis in order to improve the interpretability of coefficients. There are various advantages to doing this, for example in obtaining the standard errors for treatment groups and in the interpretation of coefficients in polynomial models, or models with interaction terms.
In the second paper Graeme Ruxton and Markus Neühauser revisit the problem of when one should use one-sided hypothesis testing. Although the difference between one-sided and two-sided hypothesis tests is taught routinely at undergraduate level, the use of one-sided tests in ecology and evolution has been controversial and debated hotly in the past. Graeme and Markus provide new insights into this debate, and provide guidelines on when one-sided tests may or may not be appropriate.