Methods Of Statistical Analysis

METHODS OF STATISTICAL ANALYSIS BY C. H. GOULDEN Senior Agricultural Sciential, Dominion Rutt Research Laboratory Honorary Lecturer in StatUtiet, Unitenity of Manitoba Winnipeg, Manitoba NEW YORK JOHN WILEY SONS, INC. LONDON CHAPMAN HALL, LIMITED IN THE REPRINTING OF THIS BOOK, THE RECOMMEN DATIONS OF THE WAR PRODUCTION BOARD HAVE BEEN OBSERVED FOR THE CONSERVATION OF PAPER AND OTHER IMPORTANT WAR MATERIALS. THE CONTENT REMAINS COMPLETE AND UNABRIDGED. COPYRIGHT, 1939 BT CYRIL H. GOULDEN Att Rights Reserved This book or any part thereof must not be reproduced in any form without the written permission of the publisher. PRINTED IN U. B. A. PREFACE From several years experience in teaching classes in statistics and giving advice at various times to experimentalists, I have come to the conclusion that there is a distinct need for more than one type of text book. On the one hand there are many who are interested only in knowing something of the theory and principles. In this class we find students who are endeavoring to obtain a broad knowledge of all sub jects related to science and art, practicing technicians such as doctors of medicine and technical advisers in agriculture, and adminislirators of research activities. It would be idle to set students of this type to work on laborious practical examples. It would probably discourage them at the start, and by absorbing time would reduce the possibility of teaching them some of the very attractive philosophical phases of the subject. In a maze of calculations the principles might be lost sight of completely, and the student would emerge with a technique for mechan ical operations and no ability to solve actual problems. At the begin ning itis not training in actual methods that is required, but the build ing up of a sound knowledge of fundamental principles. On the other hand, we have an increasing number of students who, having had some elementary training in statistics and some experience in research work, come to the point finally of requiring a practical knowledge of methods of analysis and some facility in the devices of calculation. There is no denying the fact that two or three years spent in studying the principles and theory of statistics will not fit the student to solve practical problems. To do so is to ignore the many complica tions that are involved and that training in facility is necessary in order that statistical computations may be attacked with determination and completed in a reasonable length of time. One of the objections very often raised to the use of statistical methods is the time necessary to do the routine work. Frequently this sort of thing can be Attributed to insufficient training in the actual methods that should be employed and a lack of organization of the work. The basis of this book, therefore, is the supplying of a textbook in statistics for students who have passed the elementary stage who have studied a fair amount of theory and principles and now wish to equip themselves for actual statistical work in their own field of research activities. The experiment station agronomist, the cereal chemist, the plant breeder, and the economic entomologist are all examples of research workers who require a practical knowledge of statistical methods, and undoubtedly there are many others in the same class. It has been my VI PREFACE experience that to acquire this knowledge the student must workthrough a comprehensive series of actual examples, and these should not be miniature examples as they are likely to give him a wrong impres sion of what will actually be required of him at a later date. Most of the various examples and exercises in this book are therefore of actual size, but every effort has been made to keep them within such limits as will enable the student to work through a representative set in one academic year...