[Access article in PDF]
The Voter's Guide to Election Polls, 2d edition
The Voter's Guide to Election Polls, 2d edition. By Michael W. Traugott and Paul J. Lavrakas. New York: Chatham House, 2000; pp. x + 182. $19.95.
In The Voter's Guide to Election Polls, Traugott and Lavrakas state that their goal in writing the book is to help "students of politics . . . appreciate the use of polls, especially during election campaigns" (vi). In pursuit of this goal, the authors have produced a very solid introduction to matters associated with polling. This book will be useful to all who seek to make sense of the relentless use of polling data during a campaign, be they journalists, academics, or concerned members of the public.
The structure of the book is straightforward, which helps the authors keep a clear focus. The book starts with two chapters on polls (election and otherwise), providing very nice definitions and broad overviews of different types of polls, and addressing broader process questions. The next two chapters consider the primary collectors of survey data--political organizations and media organizations. These chapters cover the reasons for data collection and how these may produce different types of data, as well as how each type of organization may utilize polling data. Importantly, the authors note potential limitations of data collected by each type of organization. The next three chapters address the nuts and bolts of data collection: sampling, interviewing, and questionnaire design. Each chapter touches on the major matters associated with the process. Next is a chapter that covers how media organizations tend to present polling results to meet the particular needs of the various types of media. The authors do a nice job of showing the limitations of the media's ability to analyze data. The last two chapters cover how a consumer can evaluate election polls and the usual litany of complaints and concerns about polls. The authors provide guidelines for how an individual can assess the value of a survey, as well as careful and full responses to the typical complaints about surveys.
With respect to content, this is a very solid book, albeit one that has significant overlap with others (such as Asher's Polling and the Public). What makes this book unique, however, is the format of the material. The authors structure each chapter as a series of questions and answers. They note that the questions were often the things asked of them by students and other interested parties. The authors use this question-and-answer format very effectively. The answers are accessible without being excessively simplified. Throughout the book, the authors provide answers that avoid technical material. There is nothing discussed in this book that requires an understanding of sampling theory or probability, or even any prior awareness of the polling process. Even while presenting material that is understandable to everyone, Traugott and Lavrakas do a nice job of covering the waterfront. For example, there is a very easy-to-understand discussion of weighting records (68-69). The authors present a simple example of how weighting works, thereby effectively illuminating what can be a rather technical topic. [End Page 742]
The unusual use of this question-and-answer format can be unsettling to the reader on first exposure. Indeed, the format of the book may limit its role in undergraduate classrooms. For classroom adoption, some instructors may simply prefer a book that has a more traditional style of presentation.
The strengths of the book include the breadth of topics covered, the accessibility of the material to the broadest audience, and the clear writing. These very strengths, however, also define the limits of the book. An instructor may want to cover matters that exceed the bounds of this book, such as a more in-depth discussion of the nature of "don't know" answers, an introduction to the specifics of data analysis, or a more technical treatment of probability. This book could very nicely serve as the primer on surveys in a public opinion course, but would...