The Total Survey Error Approach

In 1939, George Gallup's American Institute of Public Opinion published a pamphlet optimistically titled The New Science of Public Opinion Measurement. At the time, though, survey research was in its infancy, and only now, more than six decades later, can public opinion measurement be appropriately called a science, based in part on the development of the total survey error approach. Herbert F. Weisberg's handbook presents a unified method for conducting good survey research centered on the various types of errors that can occur in surveys--from measurement and nonresponse error to coverage and sampling error. Built on theoretical elements of specific disciplines such as social psychology and statistics, each chapter provides detailed treatments of the specific types of error and their potential solutions. Throughout, Weisberg is attentive to survey constraints, including time and ethical considerations, as well as controversies within the field and the effects of new technology on the survey process--in Internet surveys as well as those completed by phone, by mail, and in person. Practitioners and students will find this comprehensive guide particularly useful now that survey research has assumed a primary place in public and academic circles.