Statistics from Altmetric.com
By Antony Stewart. (Pp 160; £19.95.) Radcliffe Medical Press, 2002. ISBN 1-85775-589-8.****
According to Mark Twain “When in doubt, tell the truth”. There are few situations in medicine in which there is not an element of doubt, so we have to tell the truth. However the “truth” in medicine is never simple, especially when we give advice, as we have to function as bookmakers and call the odds: “most people with your condition survive x years”. The politically correct and intellectually challenging term for “calling the odds” is of course “statistics” and as statistics are often based on epidemiological data they are ideally presented together, and this book does so superbly (it is remarkable how many other books focus on one to the relative exclusion of the other).
This book will be invaluable for those of us whose eyes glaze over when confronted with anything but the simplest statistics. Definitions in are all in plain English. The font size is larger than in most textbooks and psychologically non-threatening, almost relaxing, suggesting “even a child could read and understand this”. Information is partitioned into short chapters; this ameliorates, but of course never abolishes, the problem of having to remember so much before one can grasp overall concepts. To assist this the author has commendably resisted further elaboration of statistical complexities and often advises that those wishing further information to consult more detailed textbooks (as listed in further reading).
Worked examples, showing how various numerical entities are derived, are particularly clear. Complex formulas (with one exception) are avoided. The concluding section features insight-provoking worked examples of practical problems that may be faced by doctors.
For most of us the message as given on page 56 is still relevant “ . . ..it is advisable to consult a statistician or someone with advanced statistical skills”. No book could replace this advice but this book offers succinct insights for those of us who, whether we like it or not, have to use or interpret statistics.
The reviewers have been asked to rate these books in terms of four items: readability, how up to date they are, accuracy and reliability, and value for money, using simple four point scales. From their opinions we have derived an overall “star” rating: * = poor, ** = reasonable, *** = good, **** = excellent.