A design where subjects are randomly assigned to treatments only within blocks.

When an observed difference is too large to believe it occurred naturally.

A sample that consists of the entire population is called a _____________.

A variable whose levels are manipulated by the experimenter.

A sampling design in which the population is divided into several subpopulations, or strata, and random samples are then drawn from each stratum.

These are hard to generate. Nevertheless, several Internet sites offer an unlimited supply of equally likely ____________________.

A component uses equally likely random digits to model simple random occurrences whose outcomes may not be equally likely.

A list of individuals from whom the sample is drawn.

A treatment known to have no effect, administered to one group to compare with other treatments.

Bias introduced to a sample when individuals can choose on their own whether to participate in the sample.

Gathering similar experimental units into blocks to isolate variability attributable to differences between blocks.

A ___________________ consists of the individuals who are conveniently available.

The entire group of individuals or instances about whom we hope to learn.