First Semester

Randomization in scientific research refers to the deliberate introduction of a stochastic element in the
design of a scientific study. Broadly defined, randomization may refer to any aspect of the study that
contains a stochastic element. For example, items in a questionnaire may have a randomized order or
the decision rule of a nonparametric test can be partially randomized to avoid the power loss resulting
from a discrete reference distribution (a so-called “randomized test”).