The Pre-Election Polls of 1948: Report to the Committee on Analysis of Pre-Election Polls and Forecasts

The Pre-Election Polls of 1948: Report to the Committee on Analysis of Pre-Election Polls and Forecasts

The Pre-Election Polls of 1948: Report to the Committee on Analysis of Pre-Election Polls and Forecasts

The Pre-Election Polls of 1948: Report to the Committee on Analysis of Pre-Election Polls and Forecasts

Excerpt

As the returns came in during the early morning hours of November 3, showing that President Truman had been elected and the Democrats had recaptured control of Congress, one of the most prominent subjects of private and public discussion was "Why did the polls go wrong?" The polling organizations were unable to give a definite answer. From many sources came a call for a detailed review of their methods and data and a prompt report to the public of whatever evidence could be obtained on the cause of the errors in the election forecasts. As explained in the Foreword, such a review was undertaken by the Committee on Analysis of Pre-election Polls and Forecasts. Its report was released on December 27, 1948, just eight weeks after the election.

The principal basis for the report was an extensive analysis of the polls prepared by the committee's staff. It is published in this bulletin substantially as it was presented to the committee on December 15. Both data and analyses are published in the belief that they will be valuable for future students of election forecasts, polling, and related phases of political and social science. They also make possible a critical examination of the evidence underlying the committee's report.

Both the committee and the staff would have preferred to devote more time to the study and to defer its publication until it had been revised more thoroughly. They would have sought to round out the investigations of certain aspects of the problem, had not other commitments limited the time which could be given to the task. It is hoped that making the data and the findings available even in incomplete and unpolished form will encourage more detailed and penetrating assessment of the election polling experience of 1948.

Under the time schedule that was adopted it was necessary to concentrate on data that could be obtained readily. The . . .

Author Advanced search

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.