Chatfield C. Problem Solving: A Statistician’s Guide

New York: Springer, 1988. — 274 p.There are numerous books on statistical theory and on specific statistical techniques, but few, if any, on problem solving. This book is written for anyone who has studied a range ofbasic statistical topics but still feels unsure about tackling real-life problems. How can reliable data be collected to answer a specific question ? What is to be done when confronted with a set of real data, perhaps·rather 'messy' and perhaps with unclear guidelines?
Problem solving is a complex process which is something of an acquired knack. This makes it tricky to teach. The situation is not helped by those textbooks which adopt a 'cook-book' approach and give the false impression that statistical techniques can be performed 'parrot-fashion'.
Part I of this book aims to clarify the general principles involved in tackling statistical problems, while Part II presents aseries of exercises to illustrate the practical problems of real data analysis. These exercises are problem-based rather than technique-oriented - an important distinction.
The book aims to develop a range of skills including a 'fee!' for data, the ability to communicate and to ask appropriate searching questions. It also demonstrates the exciting potential for simple ideas and techniques, particularly in the emphasis on the initial examination of data (or IDA). This is essentially a practical book emphasizing general ideas rather than the details of techniques, although Appendix A provides a brief, handy reference source. Nevertheless, I want to emphasize that the statistician needs to know sufficient background theory to ensure that procedures are based on a firm foundation. Fortunately many teachers already present a good balance of theory and practice. However, there is no room for complacency as theory can unfortunately be taught in a counter-productive way. For example, a student who first meets the t-test as a special case ofthe likelihood ratio test may be put off statistics for life! Theory taught in the right way should strengthen practical judgement, but, even so, the essen ce of statistics is the collection and analysis of data, and so theory must be backed up with practical experience.