NCJ Number
91858
Date Published
1982
Length
37 pages
Annotation
Following the presentation of a brief history of surveys and their uses, this paper considers the special features of large-scale surveys, problems that have concerned survey statisticians in recent years, and the present opportunities for relevant creative research.
Abstract
In concluding the review of the history of surveys, it is noted that our society has come to depend on large-scale surveys for input to policymaking, government administration, and basic and applied research. Some of the ways are used in these general areas are described. The problematic special features of large-scale surveys are identified and briefly discussed. They include high cost, being multipurpose, the period of time required for interviewing, the need for multiple interview and rotation schedules due to large sample sizes and the increasing number of surveys, and the use of multistage area probability sampling procedures. The concluding section of the paper describes a few of the major problems besetting large-scale surveys and discusses advances in understanding and controlling for their effects. The problems considered are (1) aggregate estimation and the measurement of sampling error, (2) treatment of missing data, (3) sources of nonsampling error, (4) mode of interviewing, (5) rotation group bias, and (6) statistical analysis. Seventy-six references are listed.