To make random in arrangement, especially in order to control the variables in an experiment.
see randomized clinical trial