Health Education Research, Vol. 17, No. 6, 774-775,
December 2002
© 2002 Oxford University Press
BOOK REVIEW |
Basic Statistics and Epidemiology: A Practical Guide
Antony Stewart Radcliffe Medical Press, Oxford (2002) 151 pp. ISBN 1-85775-589-8
Public Health Programme Leader and Senior Lecturer Postgraduate Medical School University of Brighton
In my more excitable moments, I would support books like these providing some sort of ennoblement for their heroic authors, a national accolade of thanks on behalf of that impoverished and deprived minority—the innumerate, the terrified ones, those who cannot mention the ‘S’ word without faint brushings of perspiration glistening on their upper lips. This book is one of those great ‘KISS’ books (Keep It Simple, Stupid); an easy, made to be understood book that even the most terrified of ‘S’ freaks can follow. (However, please see last paragraph for caveats.)
What Stewart really offers is a birds’ eye view of statistics and epidemiology; an aerial overview of the trees with occasional zooming in to inspect the wood more closely. So often, the reader or novice learner is confounded by the intricacies of specifics without getting that essential overview of the subject, which will enable them to achieve a sense of context. What Stewart achieves to some degree is to leave us with a feel for the subject, an ability to practice aspects of it with some excellent exercises and, most essentially, a ‘where do you go from here’ list of further readings.
The beginning is rather abrupt, and I felt as I read it that he was missing an opportunity here to provide those people I referred to above with an insight into the importance of statistics and what they are really all about. However, he does work within the confines of what he set out to do—a basic introduction. There is not too much scope for dallying if you plan to make those trees visible.
He really does go back to basics and his first few chapters look at those elemental aspects we really should know—how to work out proportions and percentages, how to present data, types of data (ordinal, nominal, discrete), means, medians, modes and percentiles. Then rather quickly we are onto those aspects that people often veer away from—standard deviation, standard error, how to work out confidence intervals and probabilities. It is at times like these that you realize how tenuous is the link between reading (and understanding) and remembering. However, we can glance over these lines of steps and calculations with impunity, knowing the technology exists (at a scientific calculator near you) to work them out by pressing that single, appropriate button.
I very much like the everyday clinical examples Stewart uses in this book; patients and lay people can relate to them as easily as healthcare professionals. Other mysteries such as statistical tests, what they are used for and what they actually mean are explained in real English. He does this by setting up scenarios and following them through to the point where logically you reach the ‘ahh’ stage—ahh, that’s what the test would show. He also provides useful little reminders (or potentially useful reminders if you are that way inclined) such as that 95% of all values lie within 1.96 standard deviations and 99% of all values lie within 2.58 standard deviations. You see what I mean; it is there if you want it.
Some chapters are particularly well done. Bias and confounding and screening are some in point. However, these are the sort of topics that people can relate to more easily than the ‘echte’ statistics ones. I have concentrated far more on the statistics half (well, more than half really) of the book than the epidemiology half. This is because there is not that much to say about the epidemiological chapters. They give a basic introduction, are clear, provide some indication of advantages and disadvantages of particular study types, but provide neither more nor less than half a dozen other introductory epidemiology books. I am not quite sure why Stewart did not just stick to the statistics and perhaps use the extra space to elaborate more on some of those basic concepts. I cannot help feeling it would have made a good book even better.
It is true too that basic introduction books like these can be extraordinarily frustrating. They touch on items, often based on complicated concepts, use the shorthand that is essential for their self-imposed brevity and leave the reader unenlightened. However, they are not styling themselves on the learning equivalent of ‘Eat all you can for $5’. It is recognized that they should be accompanied by a taught course that you are already studying and this simply serves to augment it; or you have already undertaken a course and this reminds you of all you have forgotten. So, for the entirely uninitiated, I would say, try it, but have a good friend you can turn to to explain it in more detail.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||