By Neil Gilbert. Price, $4.95 paper, $11.50 cloth. Pp 125. Oxford University Press, 200 Madison Ave, New York 10016, 1973.
Many clinical investigators have gained a fair understanding of and elementary facility in basic statistics. All too often such practioners fail to realize their own limitations and, from time to time, commit biostatistical malpractice. This small, personal dissertation should be perused by such investigators. It is not for the amateur. The author, who works at the Institute of Animal Ecology of the University of British Columbia, shows a sound foundation in statistical theory as well as much practical experience in biological consulting. While his arena has not been clinical investigation, this does not detract. Exposition of concepts is clear and direct; simple examples are given. There are many statements of principle and practice to which biometricians will say, "Amen!" For example, on consulting with the statistician about an experimental design, "it sometimes takes an hour to establish exactly what questions are to be asked, and five minutes to design an
Best WR. Biometrical Interpretation.. Arch Intern Med. 1973;132(4):622. doi:10.1001/archinte.1973.03650100130027