Showing papers by "Ron S. Kenett published in 1988"

IgA, IgG, IgM, and IgE levels in normal, healthy, non-atopic Israeli children.

Abstract: IgA, IgG, IgM, and IgE levels in healthy, non-atopic, Israeli-born children aged 20 days to 16 years were analyzed and showed similar age-related values and dynamics as those of white populations found in other countries No significant effect of sex of the individual or ethnic origin of the parents was found on the IgE values at different ages This may indicate that total IgE levels are strongly influenced by environmental factors Establishing tolerance limits at 975, 95, 75, 25 and 5th percentiles and the geometric mean provides the practitioner with more complete reference values The use of multivariate control charts with tolerance limits from normal IgA, IgG, IgM, and IgE levels is described and is offered as an additional tool for the diagnosis of an allergic individual

Appraisal of ceramic substrates by multivariate tolerance regions

Abstract: Determination of standards or norms and checking conformance to these standards in terms of both location and spread is a central issue in process control procedures. New observations are tested against such standards in order to determine if they are being met. In this work we apply and expand a methodology based on multivariate tolerance regions to lots of ceramic substrates used in the microelectronics industry. The method provides the appraiser with both graphical and numerical tools which include objective criteria for process control decision making. ceramic substrate plates used in the microelectronics industry. Several measurements are taken on individual substrates in order to decide whether the plates conform to an accepted standard and can be used in the production line. The plates arrive in batches and a decision-making process has to determine the conformance to standards of the individual plates and of the batches. The determination of standards or norms is a central issue in quality control procedures (e.g. Eisenhart et al., 1947; Leurgans, 1978; Fuchs & Kenett, 1987). The norms can be derived from a process capability study. The resulting sample provides a reference basis to which new observation are compared. Based on the values of all the measured variables it is decided (with prespecified margins of error) whether the new observation originates from a target reference population. The formulation of the process control problem in terms of multiple variables is very common. However, even when it is recognised that there are several dimensions the procedures applied are often univariate on key characteristics. Also common is the application of a series of univariate tests performed either simultaneously or in a stepwise fashion. The loss of information in the first case and the difficulty to assess the probability of error in the second are obvious. Another common method is to use data on all the variables to test the hypothesis of equality of means when in fact the question of interest is whether the tested observation belongs to the target population. An equivalent formulation of the latter question is to define a region covered by the target standard population and to test whether the new observation is contained in that region. Formally this defines a tolerance region for which we can assert with probability a that it contains at least a proportion P of the population of interest. The approach in this case study relies on multivariate tolerance regions with special attention devoted to the batch structure of the ceramic substrates. The product and manufacturing steps are described in the next section. Subsequent sections present the process control procedures and their application to our case study. We conclude with a discussion of the results.