Interviewer-Related Error

Interviewer-related error is a form of measurement error and includes both the bias and the variance that interviewers can contribute to the data that are gathered in face-to-face and telephone surveys. In interviewer-administered surveys, although interviewers can contribute much to the accuracy of the data that are gathered, they also can contribute much of the nonsampling error that finds its way into those data.

The methodological literature includes startling examples of measurement error due to interviewer mistakes. In 1983, an interviewer's incorrect recording of one wealthy respondent's income resulted in the erroneous report that the richest half percent of the U.S. population held 35% of the national wealth. This finding, widely publicized, was interpreted to show that Reaganomics favored the wealthy. When the error was detected ...