Accred Qual Assur (2009) 14:117–118 DOI 10.1007/s00769-008-0485-4

BOOK REVIEW

F. Ruggeri, R. S. Kenett and F. W. Faltin (eds): Encyclopedia of statistics in quality and reliability John Wiley and Sons, Chichester, UK, 2007, Four volumes (A–C, D–L, M–Q, R–Z), pp 2173, ISBN 978-0-470-01861-3, Hardback UKL 680.00 D. Brynn Hibbert

Published online: 10 January 2009 Ó Springer-Verlag 2009

With four volumes weighing in at 6.83 kg, this encyclopaedia is not light reading. It is, however, fascinating to see terms from metrology in chemistry used and explained from a different paradigm, i.e. the broader quality framework. The editors-in-chief have backgrounds in applied mathematics and statistics and come from management consulting and academia. They have collated around 400 entries written by experts throughout the world, each of four to ten pages long. In the encyclopaedia format the items are in alphabetical order, and there is a comprehensive index and cross referencing from each entry. Gerry Hahn in his forward notes the problem with the success of modern approaches to quality, ‘‘But here is the rub. Information about these tools is scattered over scores of technical journals and conference transactions, making it hard for practitioners to locate. The work, moreover, is often presented in a highly technical manner, often intelligible only to fellow researchers. This is where the Encyclopedia comes in.’’ While an assiduous reviewer might read a book of a few hundred pages, one can only dip in and out of the Encyclopaedia. There are no doubt gems that will come to light over the years. Most entries are sufficiently fundamental to expect to stand some test of time. I have therefore read this as an analytical chemist with an interest in metrology, and have tried to decide what I might want to look up. However a chemist reading the text needs to understand it was not written for chemistry per se, and how information on particular topics can be accommodated within the analytical laboratory needs to be carefully thought out. To what D. B. Hibbert (&) School of Chemistry, University of New South Wales, Sydney, Australia e-mail: [email protected]

extent, for instance, is Six-sigma appropriate for measurements with standard uncertainties in the tens of percents? The strengths of this kind of work come when reviewing a wide field of interest, such as experimental design, if they can be found. There are many entries, arising not under a section ‘experimental design’ (some references in the index), nor ‘design of experiments’ (more references in the index), but an introduction is found in ‘‘factor relationship diagrams’’ which has no cross references. This is followed, fortuitously, by ‘‘factorial designs, resolution of’’, and ‘‘factorial experiments’’ which allow further passage into the subject. The message is that what a reader wants is probably here somewhere but it may be necessary to go to and fro throughout the volumes to find it. The International Organisation for Standardisation is mentioned surprisingly little. ISO 9000 finds itself in four references in vols. 3 and 4, while ISO/IEC 17025 appears nowhere. Measurement uncertainty appears as an entry ‘‘Measurement error and uncertainty’’, which is an introduction to a series of entries on measurement systems analysis (MSA). The author observes ‘‘there are two (if not more) schools of thought with respect to the operational definitions of error and uncertainty’’. The discussion is elaborated by references to gauges, which is not so helpful to the chemist, but due respect is given to the GUM (Guide to the Expression of Uncertainty in Measurement [1]), with the Type A and Type B approach explained, how to combine standard uncertainties and a paragraph on expanded uncertainty and coverage factors. Chemistry finds its way into a discussion of proficiency testing (under the unlikely section ‘‘Multiple Modes in Proficiency Test Data’’, although PT is picked up in the index), with an example of the analysis of chromium in sediment and biota, and chlorobiphenyl in mussels.

123

118

‘‘Interlaboratory Comparisons’’ have their own entry, but without any cross reference to PT. Sampling is at the heart of much quality control activity and there is a section on ‘‘Sampling in the Chemical and Pharmaceutical Industries’’, which is linked to the article ‘‘Variables Sampling under Measurement Error’’. The presence of measurement uncertainty is clearly acknowledged in such sampling. Risk analysis is a topic not particularly familiar to many bench chemists, and yet our understanding of risk underpins the approach to quality. Related to this are ‘‘Applications of Extreme Statistics in Science and Engineering’’, an entry with 68 references. There is a leaning towards Bayesian statistical analysis, with more than a column of the index devoted to Bayes. This is entirely in keeping with modern approaches to the statistical analysis of measurement systems and fits well with the GUM approach. This seems to be a series of volumes for institutional libraries rather than the personal bookshelf, but in fact it could provide the initial go-to text for any person

123

Accred Qual Assur (2009) 14:117–118

researching or working in the field of quality. The production is good, with a nice layout and well bound pages (important as there will be much flicking backwards and forwards). Referencing and cross-referencing are variable, but it is impossible to pick up every likely reference. The format would be enhanced by electronic delivery, when some of the concerns expressed about titles of sections and indices would be solved by a decent search engine, as indeed would the weight of the four volumes be reduced. However, the journey through the pages is often interrupted by an unlikely article that catches the attention, which is the delight of paper texts.

References 1. BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML (1995) Guide to the expression of uncertainty in measurement. International Organisation for Standardisation, Geneva