The authors present a concise introduction to some epidemiologic concepts necessary to evaluate the imperfect evidence and conflicting conclusions found in the medical literature. The emphasis is on common sources of bias that plague the various study designs and strategies used in medical research.
The book is written in an easy-to-read style with a lighthearted format. Each bias is introduced as a beast whose characteristics befit the flaws and fallacies of its nature. For example, the beady-eyed "Significance Turkey," extremely impressed with the numbers he has generated and armed with several varieties of pocket calculators strapped to his belt, stands before mountains of computer printouts hawking P values.
Each chapter begins with a description of a beast and a definition of the problem to be discussed. Several case reports follow, mostly fictitious but illustrative of real problems as shown in subsequent comments and examples. Each chapter closes with further real-life