Measuring Question Sensitivity
Robin L. Kaplan
Bureau of Labor Statistics
Erica C. Yu
Bureau of Labor Statistics
Sensitive questions may lead to distortions in survey responses that create serious threats to data quality. But little research has examined if interviewers also experience feelings of sensitivity when asking respondents about sensitive topics. We aimed to understand how to measure and reduce the perceived sensitivity of survey questions from the perspective of both respondents and interviewers. Participants completed an online task where they read excerpts from fictional vignettes depicting survey interviews. We manipulated the question type (forgiving wording or direct questions) and perspective (respondent or interviewer). Participants who took on the respondent perspective rated the questions as being more sensitive than those who took on the interviewer perspective. The use of forgiving wording increased the perceived sensitivity of survey questions, and this was more pronounced for the interviewer perspective. Participants' feelings of empathy toward the vignette character and attitudes toward the survey topics also predicted sensitivity levels. We discuss implications of these results for measuring sensitivity across both respondents and interviewers.