design of experiment

"The planning of an experiment to minimize the cost of data obtained and maximize the validity range of the results. Requirements for a good experiment include clear treatment comparisons, controlled fixed and experimental variables and maximum freedom from systematic error. The experiments should adhere to the scientific principles of statistical design and analysis. Each experiment should include three parts: the experimental statement, the design and the analysis. Examples of experimental designs include: single/multi factor block, factorial, Latin square and nested arrangements.

SOURCE: apm.com"