Race of Interviewer Effects and Interviewer Clustering

26 Pages Posted: 19 Jul 2010 Last revised: 3 Sep 2010

See all articles by Janay Cody

Janay Cody

University of Notre Dame

Darren Davis

University of Notre Dame - Department of Political Science

David C. Wilson

University of Delaware - Political Science & International Relations

Date Written: 2010

Abstract

Traditional approaches to understanding race of interviewer (ROI) effects have focused more on respondent error than interviewer and survey design error. Respondents are hypothesized to alter their responses to certain items in an effort to maintain a favorable impression for the interviewer. Yet, this framework assumes respondents, interviewers, and questions all work in concert to produce the effect. The variability among these components of the survey system actually makes producing a statistically significant interviewer quite difficult. We argue that it is those interviewers who are assigned higher workloads, and therefore interact with more respondents, who are exacerbating the ROI effects to the point of statistical significance. We analyze individual and aggregate pre-election data from the 1984 National Black Election Study (NBES), finding support for our hypotheses. Feeling thermometer scores that were initially influenced by the interviewer’s race, were subsequently reduced to non-significance when interviewer workload was considered. Our findings suggest the often elusive ROI effect is potentially more related to survey design features than psychological processes related to impression management.

Keywords: Race of Interviewer, Measurement, Public Opinion

Suggested Citation

Cody, Janay and Davis, Darren and Wilson, David C., Race of Interviewer Effects and Interviewer Clustering (2010). APSA 2010 Annual Meeting Paper, Available at SSRN: https://ssrn.com/abstract=1643001

Janay Cody

University of Notre Dame ( email )

361 Mendoza College of Business
Notre Dame, IN 46556-5646
United States

Darren Davis

University of Notre Dame - Department of Political Science ( email )

217 O'Shaughnessy Hall
Notre Dame, IN 46556
United States

David C. Wilson (Contact Author)

University of Delaware - Political Science & International Relations ( email )

United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
105
Abstract Views
1,148
Rank
462,937
PlumX Metrics