TABLE OF CONTENTS Why do we sample? In some cases, we sample because we do not have the time, personnel, or money to examine an entire "population" or "lot" (all of the material of interest). In other cases, measuring a property of interest may require destroying the unit, such as in testing the life of a lightbulb or in conducting chemical tests. We may need to characterize the spatial distribution of a contaminant in soil, air, or water in an environmental situation, or we may need to characterize industrial process variation over time. We also use samples [1] for such things as process control, environmental monitoring, and product release.
What kind of sample do we want? If we want to know the "true" overall average of the lot for some characteristic, then we need a sample that is representative [2] of the lot. This means we want our sample to be a microcosm of the lot, so that whatever properties we are interested in for the lot, our sample has the same properties. For example, if we are looking for percent impurities in a chemical production batch, then we would like our sample to have the same percent impurities as the entire batch. If we are looking for toxic chemical components in soil, then we would like our sample to have the same composition as the contaminated area of interest (which may not be the entire field).
How do we get a representative sample? It's...
