Year
1977
Abstract
Statistical methods are being applied to the data base used in evaluating isotopic safeguards techniques. The statistics are used strictly as a means to achieve confirmation of the verification of the desired isotopic content. Utilizing two basic statistical approaches, paired comparisons and regression analysis, three conclusions have been reached thus far based on these statistical evaluations. These conclusions are: 1. The random variance estimates determined from paired comparison analyses (where both a reprocessing facility measurement and an independent measurement are compared) and from least-squares regression analyses, indicate that a regression model where errors in both the x and y variables are considered is desirable. 2. A regression model is needed to analyze the data from enrichment groupings which will give consistent relationships between y and x. Several models are being considered, all of which have some advantages and disadvantages. Based on a study of the methods applied to the data bank, a given model does not have a clear advantage over another model. 3. Anomalous results are identified on a more positive level using results from both paired comparison and regression analyses rather than from either one separately.