A Methodological Study of Mail Surveys

点击翻译按钮获取中文摘要

This is a request for a Small Grant for Exploratory Research (SGER) to do a study of mail survey response rates and response bias. The proposed study would be linked to an already scheduled survey planned by the Detroit Area Survey (DAS) in the spring of 1992. This study will use face-to-face interviews to collect data on racial attitudes and related issues. The proposed study would select an random sample from the DAS sample and send these respondents a mail survey where some of the questions are identical to those asked in personal interviews. It is the linkage to the DAS study that provides the rationale for requesting a SGER grant rather than a regular research grant. The proposed study will make two important contributions to the study of response rates and response bias in surveys. First it will provide a measure of the response rates that can be obtained with mail surveys compared to a highly professional face-to-face survey. This is an important issue in survey research because mail surveys are much less expensive than face- to-face interviews. The second contribution of the study will be to test the hypothesis that with self-administered questionnaires respondents are more candid about sensitive issues than they are in face-to-face interviews. The topic of the 1992 DAS study is racial attitudes--a topic that is known to be quite sensitive to the influence of social desirability factors. The proposed mail survey would include some of the same questions asked in the personal interviews thus providing a basis for comparing the answers under two different data collections scenarios. A second component of this part of the study would mail out the same questionnaire to individuals who contributed money to the recent campaign of David Duke to become governor of Louisiana. This part of the study will focus on the response rate of a sample believed to be unsympathetic to many of the issues raised in the questionnaire on racial attitudes and on a comparison of their stated views with those of the randomly selected sample.