Sunday, February 24, 2008

Sample vs. Specimen

Dictionaries are not sufficiently clear about the difference between sample and specimen, but the difference is important in the context of experimental methods. For example, we obtain a material sample froma manufacturer to determine experimentally the properties of the material. In order to do so we divide the sample into speciments, each one generating an individual data point. Statistical analysis of the data allows us to calculate such quantities as the mean value, and so on. The statistical interpretation of sample supports our adopted definition of sample (see below).

Sample (material): a piece or limited quantity of material, usually from a larger amount, taken or provided for testing, analysis, inspection, demonstration, or trial use (http://en.wikipedia.org/wiki/Sample_%28material%29). Specimens are smaller pieces taken from the sample.

Sample (statistical): the part of a statistical population which is actually observed (http://en.wikipedia.org/wiki/Sample_%28material%29); but still composed of multiple specimens. Individual items of the sample are specimens and they each generate a data point.

Specimen: An individual instance that represents a class (http://en.wiktionary.org/wiki/specimen). Not to be confused with a sample, from which specimens are obtained.

No comments:

Post a Comment