Some Statistical Perspectives for Research and Development

This book is aimed at those who have most likely completed an introductory unit or units on statistical methods and who now wish to intelligently use statistical techniques in their research or development work. It is not discipline specific but has a bias towards those in the physical or biological sciences as well as engineering, although those working in other areas are still likely to find the content valuable. Basic mathematical and statistical material is covered but is done so in a way that requires a certain maturity of the reader, as the book does not contain graded exercises nor additional support material which are the norm for course texts. The book is, therefore, unlikely to be considered for a course text but rather as support material for undergraduate or postgraduate units on statistical methods or as a shelf reference. It is also designed to be self-instructional for highly motivated readers. All the techniques presented are illustrated by worked examples which also demonstrate the use of appropriate statistical software. The monograph covers a greater range of topics than is usual for a single volume on statistical techniques. Standard single variable methods are covered with a short chapter on Clinical Trials that, amongst other things, briefly discusses some of the prevailing ethical issues. There are additional chapters on Bootstrapping and Sampling Finite Populations and the Role of Statistical Techniques in Quality Improvement. This latter chapter has been written specifically with those in mind who have or will be working on industrial development projects in this area. Besides these, material is presented on Statistical Design of Experiments, Multiple Regression and Multivariate Statistical Techniques. Throughout, constant reference is made to the problem of determining sample size and attempts are made, in each instance, to resolve this matter as it is invariably a burning issue amongst researchers yet a topic that is frequently ignored by many statistical texts. No apologies are made for dealing with theoretical issues under-pinning the techniques presented before discussing some of the more practical matters that invariably arise. It is acknowledged that readers may frequently need more advanced material on certain of the topics than have been covered and so numerous pointers to additional reading are given.