Obtaining confidence intervals around any measure is currently a strong trend for the measurement laboratories. Concerning tensile testing, this trend is embodied into the UNCERT group that issued the Codes of Practice. The procedures of the measurement laboratories have to be checked for conformance to these requirements.
We have developed and validated a new procedure that computes the limits of the pertinence interval of the linear model when determinating the Young's modulus. The key point of this procedure is the choice of an efficient quality factor to rank the possible choices for this interval.
By the way, a careful analysis of the collected data has been undertaken, and it has been shown that these data are not a simple mix of "few mechanical constants'' and "a lot of noise''. Indeed, it appears that this noise was not only carrying information in the Shanon's acceptance of this term, but also carrying quanta of effective knowledge.
The data collected for a lot of test pieces can be used to characterize not only the test pieces themselves, but also the testing testing machines and the testing process. To achieve such a goal, the use of data mining methods is required.
Moreover, it has been shown that many properties can be derived from the fact that, due to discretization, the collected data are not random points in a continuum, but rather integer multiples of some quantum.