Randomized

From EvaluationWiki

In the evaluation context, considerable importance is attached to research designs in which the allocation of subjects between the experimental and control groups is randomized (see RCT). It is therefore noted here that many common methods of doing this are not acceptable, e.g., by flipping a coin. Many such methods, although intuitively chance-based, have been shown to give non-random results in the relevant sense. For example, flipping a coin will inevitably result in a string of ten heads, sooner or later; if you are running a small number of subjects, this will obviously not provide a way to allocate subjects. But almost all mathematical functions for producing random numbers actually produce pseudorandom sequences; wikipedia.org has a good article on pseudorandom numbers.