The Fraction of Missing Information as a Tool for Monitoring the Quality of Survey Data

摘要

The response rate has been a longstanding indicator of the quality of survey data. Recent empirical research, however, has called into question the value of the response rate as a proxy indicator for nonresponse bias, which is the most damaging aspect of nonresponse. A new measure for the risk of nonresponse—the fraction of missing information (FMI)—is proposed as an alternative to the response rate. FMI was developed as part of the methods for handling missing data, and it measures the level of uncertainty about the values one would impute for current nonresponders. Developed as a tool for monitoring ongoing data collection, FMI can help researchers maximize the information in the final dataset. This article illustrates its application to telephone and face-to-face surveys.