REVIEW ARTICLE/EDUCATIONAL PANEL
How to plan an experiment I. Randomization: current fad or (ever)lasting fashion?

All those who like experimentation are sooner or later faced with a need of random selection of elements or objects of interest that they want to study (persons, patients, animals, cells, etc.). Randomization, a basic requirement in appropriate planning of experiment, may be performed either to select series of randomly assigned elements/objects or to allocate the studied objects to a given group, medical or diagnostic procedure, treatment protocol, etc. The principal advantage of random selection is to minimize effects of bias and confounding variables, two fundamental threats known to weaken research credibility. Simple, unrestricted random selection can lead to undesirable imbalance in baseline characteristics, thus affecting any credible reasoning. Restricted randomization (with blocking or stratification) includes procedures used along with random sampling that help to achieve balance between study groups in their baseline characteristics or in size. For all researchers conducting biomedical studies randomization ensures a straightforward and reliable analysis of the outcomes, and enables any further generalization of their findings.