Introduction to Medical Biometry and Statistics

by Raymond Pearl '99, W. B. Saunders Philadelphia.

Any one who knows anything about the author of this book and some of his work knows at once that whatever he undertakes will be thoroughly and well done and will be reliable. Under the circumstances the reviewer is especially pleased to find such an authority emphasize a fact which is overlooked very frequently by those who are relatively unfamiliar with practical conditions which form the basis of this fact. This fact may be stated briefly, that many applications of relatively elementary mathematics are going to. be made by biologists, medical men, psychologists, educators, etc., whether mathematicians like it or not. The reviewer would go further and say that the mathematician who disdains to give his attention to these applications simply because the mathematical principles are relatively elementary and refuses to help to set up danger signals in dangerous spots should not be surprised or too cynical if erroneous results are obtained occasionally by those who lack a thorough training in mathematics. In fact, the mathematician is very much to blame because of his failure to assume his part of the responsibility.

Professor Pearl has done a remarkably good job in presenting statistical theory to biologists and medical men who' have had a minimum amount of training in mathematics and he properly emphasizes the fact that such men are first of all biologists or medical men and are therefore just the individuals who are most responsible for the proper interpretation of the data. The reviewer feels, however, that it is rather dangerous to encourage such individuals to lose any of their respect for the fundamental principles of mathematics and that any investigation in which this - respect is lacking is fraught with difficulties and pitfalls. However, the reviewer is not sure enough that Professor Pearl does leave this impression to ofifer this statement as a criticism- Much would depend upon the way a course based upon the book as a textbook was taught- The book is not intended to be used and could not very well be used as a textbook in a course in mathematics, for much of the book is given over to a practical treatment and discussion of what he calls biostatistics and includes a thorough discussion of medical records and their mechanical tabulation, in addition to the usual treatment of tabular presentation of statistical data and the graphical representation of these data. The latter part of the book treats of probability, the probable error concept, variation, correlation theory, curve fitting, etc. One feature of the book is a chapter devoted to a short but interesting history of the more recent development of statistical theory. The book meets remarkably well the needs of the medical man who is interested in the application of statistical methods in his field.