Health Education Research, Vol. 17, No. 2, 274-275,
April 2002
© 2002 Oxford University Press
BOOK REVIEW |
Making Sense of Data: A Self-instruction Manual on the Interpretation of Epidemiological Data, 3rd edn
J. H. Abramson and Z. H. Abramson Oxford University Press, Oxford, 2001 367 pp, ISBN 0195145259 (pb)
Postgraduate Medical School University of Brighton
This superb book is a real boost to anyone attempting to interpret data from studies accurately. It provides the necessary tools, aided and abetted by the reader's common sense, to allow for a systematic and informed approach to interpreting and using data on which to base clinical and healthcare decisions.
Inaccurate and inappropriate interpretation of data is a major hindrance to effective clinical decision making. This book aids the reader in acquiring those skills which would allow him or her to differentiate data that is useful from that which is biased, the rigorous from the tentative and the usable from the plain rubbish.
When I first started reading the book I was confused by the way the exercises and answers had been organized as answers were in a subsequent unit to the questions. I'm not quite sure why the authors decided to organize it as such, but once the system is fathomed it becomes an insignificant issue. Indeed, the system is vastly removed from the old-fashioned one whereby a set of questions and answers co-exist in an organized fashion, but the expense of context. The way the Abramsons have dealt with this problem is to integrate the answers into the discussion rather than provide sets of isolated exercises. This incorporation of answers into a single narrative provides the book with an excellent flowing style.
Having criticized organization, I find myself using the same word to describe one of its great strengths. There are many strengths in the book, but perhaps the way it is organized is the most creative. The authors have made use of a clear numbering system to connect related exercises, discussions, answers to the exercises and notes to build discrete, but connected, learning concepts. Readers are thus able to spend as much or as little as they individually require on each area being developed. The notes that the authors have incorporated into the sections are used to clarify points, develop them further by introducing more advanced knowledge (e.g. confounding on 40 ) or to identify additional reading and to provide references.
What I also like about this book are the touches of humour that are introduced, e.g. when considering appraising the validity of a measure, the authors use a fictional disease, ‘TV dementia...caused by excessive exposure to television’ to prevent the reader ‘being influenced by your prior knowledge about the measure’.
I mentioned context earlier and contextualization is a philosophy to which the Abramsons clearly adhere. Using examples, diagrams and stories that the readers can visualize, the authors place the concepts into the real world, rather than a world of numbers and abstractions. In fact, this device of visualization is enhanced by their use of asking and answering questions, and the problem or scenario based method of delivery. In addition, examples they use are interesting and simple, rather than simplistic. They seem to work on the principle of a ‘need-to-know’ basis rather than challenging the reader's arithmetic or mathematic abilities.
Context is also the subtext for ensuring the wood is not lost for the proverbial trees. The device they use to ensure that we, the readers, understand the purpose behind the techniques is to constantly, but not repetitively, go over the tools that have been acquired, focus on smaller pieces of the puzzle, and then invite the reader to step back and view the picture that has emerged in its entirety. It would be easy to get lost in a maze of intricate exercises, but the authors provide a cross-reference of where to find information within the book for revision. They never allow us to lose sight of the underlying reason for doing the exercises; to ensure we have the basic procedures for appraising data. This includes asking the elementary questions:
- What are the facts?
- What are the possible explanations?
- What additional information is required?
These three questions, deceptive in their simplicity, form the basis for being able to understand and interpret data or information, whilst at the same time enabling that crucial commonsense overview to underpin any decisions made.
The book is divided into seven chapters starting with an introductory chapter on basic concepts and procedures, and moving through rates, measures, associations, and causes and effects, to a gem of a chapter on meta-analysis (the critical review and integration of the findings from separate studies). This penultimate chapter provides an overview of the contentious, yet potentially invaluable, methodology and statistical technique of meta-analysis. It includes a clear explanation of heterogeneity and an appropriate warning to bring differences to the surface and examine them, `rather than drowning them in a statistical pool'.
Apart from using this book as a textbook from which to learn basic concepts and principles, it is an invaluable reference source. I would strongly recommend this book not only to students or teachers of epidemiology, teachers of appraisal who would have a field day plundering the rich source of examples (and answers—always an added bonus), but for anyone who wishes to have a better understanding of how to appraise and use data in daily clinical decision making.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||