U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Survey Design and Analysis - Current Issues

NCJ Number
74003
Editor(s)
D F Alwin
Date Published
1977
Length
154 pages
Annotation
This collection of papers examines current issues in survey research methodology, with emphasis on sources of errors and strategies for dealing with errors.
Abstract
An introductory paper develops a framework for considering issues in survey methodology, especially those concerning errors in surveys, and presents an overview of research in the area. Topics discussed include completion rates, sample coverage, locating respondents in longitudinal research, and response rates. Other topics include item nonresponse, weighting to adjust for noncoverage and nonresponse bias, interviewer variability, question structure and sequence, methods of administration, and respondent errors. A paper on the wording of questions concludes that the question form frequently affects the results. An examination of the effects of interviewers' prior expectations on responses to threatening questions concludes that in most practical situations, these effects are trivial and can be ignored. A study of the impact of informed consent on response rate and response bias indicates that the use of voluntary participation and consent reduces participation but does not introduce bias into the three criterion measures studied. Also, an examination of 38 different social surveys indicates that patterns of response to mail surveys fit the gamma distribution; this finding permits the development of effective call-back procedures to nonrespondents as well as minimization of the waiting time for completion of a survey. Another paper presents procedures for treating missing data in the statistical analysis of survey data. A final paper examines response error in earnings functions for nonblack males and concludes that logarithmic earnings is one of the most accurately measured indicators of socioeconomic success. For most papers, figures, tables, and notes are included. Reference lists are provided.