Consideration of statistical aspects is a priori indispensable when planning an experiment. Following the formulation of the null hypotheses, the strategy for evaluation and the statistical test methods to be applied have to be decided upon. The level of significance, α, the power of the statistical test and the "effect size" have to be considered when calculating the sample size.
Quality. and validity of any statistical test can be sized up properly only if this information is cited when research results are published. Lack of this information makes it impossible to verify whether results were obtained in a methodologically sound context, or may lead to wrong conclusions because of a in most cases insufficient sample size.
Using t-test, analysis of variance, and 2 x 2-contingency table for illustration, methods for determining sample size and power of statistical tests are described and elucidated by examples.
A literature review provides an impression of sample sizes for experiments in forest research. Unfortunatelv. there has been, as a rule, not enough information provided concerning statistical testing; thus it is almost never possible to judge the validity of results from a statistical point of view. In order to assure the comparability of scientific studies, statistical methods used should be described in detail both when planning the experiment and also when publishing the results. Especially needed is information on the used statistical test, the power of the test, the number of replications, the effect size and the level of significance, α, i.e. the exact p-values should be presented.