4/14/2014 feature story
Trivellore Raghunathan explores how measurement error and nonresponse bias change as a result of survey length. In particular, he compares the impact of using a subset of questions from a survey along with imputation for omitted questions to using the full survey.
Key indicators used for policy making often come from large-sample surveys. The growing need for such data has resulted in longer surveys, and in some cases, substantial burden for respondents. Evidence indicates that survey length affects participation, leading to greater nonresponse and increased potential for nonresponse bias. Longer surveys have also been linked to suboptimal responding, resulting in measurement error. Thus, long interviews can bias survey estimates and lead to misinformed decisions. This project estimates the effects of nonresponse and measurement error in both interviewer- and self-administered modes of data collection. It also implements a split questionnaire design, randomly assigning respondents to receive a subset of the survey questions, and uses multiply-imputed analysis for the omitted questions. Our main hypothesis is that the subset approach will yield estimates with less bias and even less total error than the full questionnaire approach. To evaluate the extent of the problem and implement a solution, this study will: 1. Examine whether measurement error increases as a function of survey length; 2. Isolate the impact of survey length on nonresponse bias; and 3. Evaluate the reduction of bias and impact on mean square error from using split questionnaire design, after multiply imputing the full data for all respondents. The long-term objective of this study is to provide empirical evidence leading to a paradigm shift in survey design that will allow for improved information on health while reducing respondent burden.
- News & Events
- Info Services