Abstract
This study investigates how an onscreen virtual agent's dialog capability and facial animation affect survey respondents' comprehension and engagement in “face-to-face” interviews, using questions from US government surveys whose results have far-reaching impact on national policies. In the study, 73 laboratory participants were randomly assigned to respond in one of four interviewing conditions, in which the virtual agent had either high or low dialog capability (implemented through Wizard of Oz) and high or low facial animation, based on motion capture from a human interviewer. Respondents, whose faces were visible to the Wizard (and videorecorded) during the interviews, answered 12 questions about housing, employment, and purchases on the basis of fictional scenarios designed to allow measurement of comprehension accuracy, defined as the fit between responses and US government definitions. Respondents answered more accurately with the high-dialog-capability agents, requesting clarification more often particularly for ambiguous scenarios; and they generally treated the high-dialog-capability interviewers more socially, looking at the interviewer more and judging high-dialog-capability agents as more personal and less distant. Greater interviewer facial animation did not affect response accuracy, but it led to more displays of engagement—acknowledgments (verbal and visual) and smiles—and to the virtual interviewer's being rated as less natural. The pattern of results suggests that a virtual agent's dialog capability and facial animation differently affect survey respondents' experience of interviews, behavioral displays, and comprehension, and thus the accuracy of their responses. The pattern of results also suggests design considerations for building survey interviewing agents, which may differ depending on the kinds of survey questions (sensitive or not) that are asked.
Highlights
An important source of knowledge about society is what people report in survey interviews that produce the data for official statistics, e.g., population estimates on employment, health and crime
The current findings extend work on people’s reactions and behaviors when they talk with interviewing agents, for example telling stories to an agent that exhibits listening behavior (e.g., Gratch et al, 2006; von der Pütten et al, 2010), answering openended questions asked by a peer (e.g., Bailenson et al, 2006) or answering open-ended questions asked by a deception-detecting kiosk agent (Nunamaker et al, 2011), to the task of a survey interview for social measurement that uses closed categories as response options and that is designed to make statistical estimates of a population
The findings extend work on disclosure of sensitive information in a survey interview with a virtual interviewer (Lind et al, 2013) to an interview with non-sensitive questions that have verifiably correct and incorrect answers, and in which accurate comprehension of the terms in the questions is critical
Summary
An important source of knowledge about society is what people report in survey interviews that produce the data for official (government) statistics, e.g., population estimates on employment, health and crime. In the most strictly standardized interviews, interviewers are required to ask questions exactly as scripted and use only “neutral probes” like “Let me repeat the question” or “Whatever it means to you” if respondents say anything that isn’t an acceptable answer (e.g., something other than a response option included in the question), so as to ensure that all respondents receive the same stimulus and to avoid the possibility that interviewers will bias responses (Fowler and Mangione, 1990) This can lead to perverse interactions in which interviewers thwart respondents’ efforts to understand what they are being asked by refusing to provide the clarification that respondents seek (Suchman and Jordan, 1990), and in which interviewers violate ordinary norms of conversation by failing to “ground” the meaning of utterances they themselves have produced (Schober and Conrad, 2002). The best response accuracy, overall, seems to result when respondents request clarification if they believe they need it (“What do you mean by work for pay exactly?”), but when interviewers can volunteer clarification when they believe respondents need it (Schober et al, 2004)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
