Thoughts on quality systems

An intrinsic, universal definition of quality has been quite elusive. As applied to products and services, the impressions of users are so subjective as to escape definition. To some extent producers see it in terms of what people will continue to buy, and if they will buy large amounts, the quality must be exceptionally good. While an absolute scale may escape us, changes along the scale do not. If a product (say coffee) or a service (say a haircut) changes by even a small amount we tend to notice and judge it as being higher or lower in quality. Quality systems, for example ISO 9000, and ISO 17025, can be used to facilitate products and services that are acceptable to clients by avoiding decreases in perceived quality. More about this in a moment. There is a physical approach to defining quality that is based on matching or exceeding defined physical requirements. This is an arena where formal quality systems have great utility. The client can test to see if his requirements are indeed met. Yet, even here the definition of quality may elude us. When the components are assembled into products, they will be judged by the customers based on intangibles that go beyond individual components exceeding requirements. Overall design, color, feel, and user friendliness will have a role in the customers’ judgment of quality. Even that judgment is likely to be relative to other products (or the ‘‘last one I owned’’) without any quantitative figure of merit on an absolute scale. Taking all of this into account, we see that there is not yet an acceptable scale for measurement of quality. Lacking that scale, we cannot quantify even small changes up or down from a fixed point. However, we usually can judge if a different product or service is better or worse than the previous one with respect to whatever attributes of quality we choose. This restricts any attempt at formal quantification. The way quality systems serve to assure meeting defined physical requirements is essentially the same as for meeting customer perceptions: that is to reduce variability to an acceptable or specified minimum. Reducing variability (improving the precision of the process) is critical before the process can be ‘‘tweaked’’ to improve quality. Without a stable system there can be no meaningful adjustments. A given adjustment in an unstable system may provide an instantaneous improvement in perceived quality, but there can be no assurance that the improvement will be reproducible. Thus we see quality systems are of utility in applications where rigidity and invariance are required to maintain acceptably low variation in output. Overall, this situation is found in chemical laboratories only where there is an ongoing pattern of analyzing very similar items. Examples could include clinical, steelmaking, sports doping, and soil testing labs. The emphasis in all these is routine, with sufficient repetition to assess variability and adequately control it through exercise of rigid procedures such as those developed to comply with ISO 17025. Quality systems are likely to be less helpful in chemical laboratories where analyses are not routine. That could include academic, high-level metrology, investigative, and consulting labs where very diverse materials are analyzed. This observation may help us to answer a question raised by such studies as IMEP. (Details on all Interlaboratory Comparisons of IMEP are available at: http://irmm.jrc.ec. europa.eu/html/interlaboratory_comparisons/imep/index. htm). Expert analytical chemists recognize IMEP for its S. D. Rasberry (&) 1255 Glebe Road, Lottsburg, VA 22511, USA e-mail: rasberry@rivnet.net