ABSTRACT
Much of the data that statisticians gather to gain information can be thought of as a sample from some larger collection of objects — which they will refer to as the target population. A sample from some population is said to be representative for some attribute of the population if the proportion of objects in the sample with this attribute is close to the population proportion with this attribute. A random sample from some population is highly likely to be representative for any given attribute of this population, provided its size is sufficiently large. Minitab is capable of obtaining random samples, both with and without replacement, from a finite population. It does this by sampling from the set of nonempty rows in a designated column, and putting the sample in some designated column. When the population median is not unique, the behavior of the sample median tends to bounce around, never really settling down.
