A study of the impact of bias on accountability

In the area of production the primary goal is to produce a product that meets (or exceeds) customer expectations. This requires that one first obtains statistical control of the process at some desired level of operation. Once this is accomplished, one looks for ways to reduce variations around that level. The authors take corrective actions only when statistical control charts indicate that some undesirable change has occurred. They do not, for example, concern themselves with the possibility of relatively small changes that routine control chart tests do not detect. Relatively small biases are of no great concern. However, significant deviations from the desired level of operation are of great concern. In the area of materials accountability the situation is somewhat different. Here the primary goal is to maintain the best possible records of material inventories taking into account the measuring uncertainties that are associated with such inventories. In addition to the variations of individual values around averages (precision), the authors must also concern themselves with those deviations in averages from their true values (biases). The authors are faced with the problem of knowing true values with exactness, and they are faced with the problem of needing to identify those relatively small biases that behave as random variables while exhibiting little or no statistical significance