Measurement Errors in Reported Race-Related Attitudes by Race of Interviewer, Perceived Race of Interviewer, and Race of Respondent

• Activity Number: 531 - Survey Modes and Measurement Error
• Type: Contributed
• Date/Time: Wednesday, August 1, 2018 : 10:30 AM to 12:20 PM
• Sponsor: Survey Research Methods Section
• Author(s): Paul Lavrakas* and Dan Thaler and Lin Stork and Del Solis
• Companies: Self-Employed - Independent Consultant and Michigan State U. Office for Survey Research and Michigan State U. Office for Survey Research and Michigan State U. Office for Survey Research
• Keywords: Race-related; Attitudes; Measurment Bias

Abstract

A 2017 Pew study reported that respondents were not very accurate in identifying the race of the interviewer with whom they had just completed a telephone questionnaire. Similar to the Pew methodology, we conducted a state of Michigan DFRDD survey in 2016 in which each respondent was last asked "Finally, what do you think is my racial or ethnic background?" Our findings indicate that Whites and Blacks were (1) were much more accurate in identifying the race of White interviewers than of Black interviewers; and (2) that Blacks were more accurate in identifying Black interviewers than were Whites. More importantly we also found that the reported answers about the race of Black interviewers given by White respondents was significantly related to the substantive answers they gave to racial attitudinal questions. In addition to explaining that measuring race related attitudes accurately in an interviewer-related survey is even more complex than previously thought, we also will explain why the Pew approach and our approach to measuring the ability of the respondent to correctly identify the race of the interviewer is fraught with social desirability bias.