General Interviewer Techniques: Developing Evidence-Based Practices for Standardized Interviewing

Authors

Nora Cate Schaeffer, University of Wisconsin Survey Center, University of Wisconsin-Madison; Department of Sociology, University of Wisconsin-MadisonFollow
Jennifer Dykema, University of Wisconsin Survey Center, University of Wisconsin-MadisonFollow
Steve M. Coombs, University of Wisconsin Survey Center, University of Wisconsin-MadisonFollow
Rob K. Schultz, University of Wisconsin Survey Center, University of Wisconsin-MadisonFollow
Lisa Holland, , Survey Research Center, University of MichiganFollow
Margaret Hudson, Survey Research Center, University of MichiganFollow

Date of this Version

2-28-2019

Document Type

Article

Citation

Presented at “Interviewers and Their Effects from a Total Survey Error Perspective Workshop,” University of Nebraska-Lincoln, February 26-28, 2019.

Comments

Copyright 2019 by the authors.

Abstract

The practices of standardized interviewing developed at many research sites over many years. The version of standardization that Fowler and Mangione codified in Standardized Survey Interviewing has provided researchers a core resource to use in training and supervising standardized interviewers. In recent decades, however, the accumulation of recordings and transcripts of interviews makes it possible to re-visit the practices of standardization to describe both how respondents actually answer survey questions and how interviewers actually respond.

To update General Interviewer Training (GIT), we brought observations of interaction during interviews together with research about conversational practices from conversation analysis, psychology, and other sources. Using our analysis of the question-answer sequence, we identified the principal actions covered in training as reading a survey question, recognizing a codable answer, acknowledging a codable answer, and follow-up for an uncodable answer. Our analysis of each of these actions is influenced by our observations of the participants’ behavior – interviewers must be trained how to repair the reading of the question, for example -- and by how that behavior is influenced by characteristics of survey questions – follow-up differs for yes-no and selection questions. We developed a set of criteria to use in evaluating the likely impact of the choices we recommend on, for example, interviewer variance and the motivation of the respondent. Although research is not available for all (or even most) criteria, we attempted to be systematic in assessing the likely costs and benefits of our decisions.

We focus on standardized interviewing, which attempts to train interviewers in behaviors that all interviewers can perform in the same way. However, the evidence supplied by studies of interviewer-respondent interaction makes clear that the impact of the question on the respondent’s answer, and the way that respondents answer questions must be taken into account in any style of interviewing.