ABSTRACT
In the quality statistical world, quite often, we are interested in tests and condence intervals for population means based on random samples from those populations. It is a common inquisition for most of us in the eld. To make things easier for the experimenter let us begin by identifying the terms and their meaning. As used quite often, “test” means a test of a hypothesis, set up prior to the experiment (a priori), that the population mean has a specic numerical value. The test yields simply the conclusion that the data are or are not consistent with the hypothesis. The method of testing depends on whether the population standard deviation is known or unknown. In experimentation, the case of unknown standard deviation is the more important and usual; in repetitive work, such as production quality control and standard tests, the standard deviation may sometimes be calculated closely from the large amount of earlier data, including surrogate data.
