Folks: This is off topic, but I believe many R-Help participants would be interested in this. My apologies to my British colleagues, who probably already know about this, and to others for whom this is a waste of their time. Dr. Ben Goldacre, a British Physician and science columnist, has written a book for popular consumption entitled: " Bad Science: Quacks, Hacks, and Big Pharma Flacks" . While much of it is concerned with fraudulent products (Homeopathic treatments, nutritional supplements that reduce wrinkles,grow hair, etc.), readers may find large parts of it relevant to the role of statistical thinking in science (mostly medicine). For example, we all know about the placebo effect, but I found his chapter on it fascinating. He also has a lot to say about publication bias and why the public at large -- especially journalists -- need to understand the role of statistics in evaluating scientific claims. His website and columns may also be of interest: http://www.badscience.net/ Cheers, Bert -- Bert Gunter Genentech Nonclinical Biostatistics