THIS book aims to present as many opinion poll results as possible, in a convenient and useful form. The material presented has been collected from 23 organizations in 16 countries and covers the period from 1935, when George Gallup and Elmo Roper began publishing results obtained by the sampling method, through 1946. It is hoped that subsequent volumes covering five-year periods may be compiled.
Although we have included most questions asked of national cross sections throughout the world, this collection by no means includes every question ever asked in a public opinion poll. Market research material was not available, and because of space limitations we have been unable to incorporate results from local polls.
The plan of the book is a simple one. In general, the classification and wording of both subjects and cross references are based on the Library of Congress subject headings which are the result of careful thought and wide experience. To insure uniformity, it seemed both wise and economical to adopt a system that has long been tested in practice through its wide use in libraries. Each question has been classified under the most specific subject that could be assigned to it; under each subject, questions have been arranged in chronological order except where questions on a single topic have been assembled for presentation in tabular form, when the entire series appears under the date of the earliest question.
Many questions involve two, three, and sometimes more ideas or subjects. In these cases, the choice of subject had to be arbitrary, but cross references have been provided in the Table of Contents to connect such questions with all the subjects involved. To retain the sense of filter questions, it seemed essential that they be kept with the questions that follow and we have followed this practice since limitations of space prevented the publication of any question more than once.
In some cases, questions on successive phases of a single topic have been classified under different heads, for example "International organization" and "United Nations"; "World War, 1939-1945: Food questions" and "Food relief". In the case of the United Nations, the earliest questions are classified under "International organization". A break was made with January 9, 1942, when the organization was given its present name—all questions after that date are under "United Nations". Similarly, questions asked during the war about food relief to Europe, etc., are classified under "World War, 1939-1945: Food question", and questions asked after the war are under "Food relief". These are arbitrary divisions, really, calculated partly to keep the size of the subject down to reasonable proportions, and partly to separate the successive phases of a subject.
Cross references refer to questions and not to answers. To have referred also to answers would have been impracticable. The subjects themselves are printed in capitals in the Table of Contents; all other entries are cross references.
We have made few exceptions to the rule of using only results based on a national cross section. The first survey made in France after the Allies took Paris was made only in Paris for lack of travel facilities. In the Netherlands, the first survey covers only the three western provinces—the rest of the country was still in the hands of the enemy. In Brazil, because of the peculiarities of the country, the two large cities— Rio de Janeiro and São Paulo — are the best cross section obtainable. A considerable number of surveys in Hungary were only of Budapest and suburbs because of interviewing difficulties in the rest of the country.