ACSPRI Conferences, RC33 Eighth International Conference on Social Science Methodology

Font Size:  Small  Medium  Large

Verbal and non-verbal paradata of income questions: indicators for response quality and question comprehension of respondents

Marieke Haan, Yfke Ongena

Building: Law Building
Room: Breakout 8 - Law Building, Room 100
Date: 2012-07-11 01:30 PM – 03:00 PM
Last modified: 2011-12-17


Paralinguistic and visual displays are an important element of communication in survey interviews (Conrad & Schober 2007). Paralinguistic displays comprise vocal cues, such as pitch, intonation, speech rate, fluency and loudness (Barath & Cannell, 1976; Van der Vaart et al., 2006). Visual displays are nonverbal behaviors not directly related to speech but they are affected by linguistic aspects. Examples of these displays are gaze, gesture, facial expressions like smiling or lifting an eyebrow and nods (McGovern & Tinsley, 1978; Conrad et al., 2008).
The experiments that have been conducted in this field found that gaze aversion indicates that the respondent finds the question difficult (Doherty-Sneddon et al., 2002) and that speech disfluencies can be a sign that the respondent needs clarification (Schober & Bloom, 2004). Jans (2010) also showed that verbal paradata as voice pitch and disfluencies can predict the response on income questions given by respondents.
In this study the effect of verbal as well as non-verbal paradata on the interaction between the interviewer and the respondent in income questions is further investigated, since non-response of income questions is a problem when studying survey data (Jans, 2010). Video recorded face-to-face interviews have been coded in Sequence Viewer (Dijkstra, 2002). It is examined if the verbal and non-verbal cues given by the respondent can give information to the interviewer about the comprehension of the question and the quality of the answer of the respondent. In addition to this it is studied if the paradata given by the interviewer influences the responses of the respondent. It can be expected that disfluencies of interviewers affect the respondent’s question comprehension and that making eye contact with the respondent affects the response quality of income questions.