|
A criterion and evaluating method of agreement for the determined and true qality of particulate mat |
|
|
|
|
A criterion and evaluating method of agreement for the determined and true qality of particulate material
Author: J. Martyniak
In this paper there are sayed the significantest terms pertaining to the research into closeness of agreement between the true characteristics of particulate material lot and the analysis results correspond with them. A convenient information on this closeness is the difference between the independent replicate estimates representing the same product property. This difference indicates the precision level of the used estimation proeedure that forms the goodness of approximation to the lot true values for determination results. Therefore, an adequately criterion of equality for the true and determined value of a lot characteristic is the required diffcerence limit for two replicate estimates. This limit is termed the permissible difference and is a measure of test reproducibility has been attained during the estimation procedure. If the differenee between replicate estimates did not exceed such a limit, the used estimation procedure had been fited precision then it is assumed that the estimation result and the true value of property of particulate product lot are equal. There are presented the essentials to shape and calculate the required difference limit. It is taken into consideration the circumstance that the circumscribe of product quality most frequently is multidimensional or multivectorial and the applied estimation procedure is a multistage with a various number of stages for diverse properties of product. The principles of the multidimensional and multivectoriaI estimation are the fundamental to advance and show the mathematical model of statistical closeness fits for every characteristic set. The point of departure to solve the considered problem is detecting and evaluating all the characteristic deviations arise from succesive stages of estimation procedure. These deviations are the basis to be computed the cumulative deviations of estimates after every stage in relation to the true product quality parameters.
|
2011