- We may be interested in a
*unit*because we intend to generalize from observations on a few units to a statistical assembly of units. Opinion polling and survey sampling provide well known examples of this type of research. - We may be interested in the dynamics of a
*unit*, how its observable characteristics change from time to time. Economic studies of business firms provide an example of this type of research. (See dynamic model.) - We may be interested in the internal functioning of a
*unit*which we can characterize as a statistical assembly. This kind of research often involves interference with the unit such as subjecting it to a treatment or even dissection, in some cases. Field experimentation and clinical trials. are examples.