Does Interviewers’ Age Affect Their Assessment of Respondents’ Understanding of Survey Questions? Evidence From the European Social Survey

Abstract

General population surveys target the entire adult populations of participating countries, where, in many advanced societies, older respondents constitute an ever increasing share. Reflecting these processes, in the European Social Survey (ESS), the proportion of older respondents has been increasing. 2 Prior research has found that older respondents are more often assessed by interviewers as having problems with understanding survey questions. This has been attributed to older respondents' lower cognitive performance resulting in lower quality responses. While researchers recognize the social character of the face-to-face interview situation, till now, analyses of the links between age and data quality have focused on characteristics of the respondent while largely ignoring the characteristics and active role of the interviewer. One possible reason for this is that data on interviewers' characteristics in many publicly available datasets are either lacking or very limited. Ignoring the role of the interviewer may also be related to the wider tendency to treat survey (quantitative, in general) data as "objective", despite methodological evidence to the contrary.

What can Interviewer-Collected Paradata Tell About Measurement Quality in Face-to-Face Surveys? Analyzing Response Styles in six Rounds of the European Social Survey

“…At the same time, a respondent's motivation and understanding were associated negatively with SL and ERS, and, in the case of motivation only, with a middle response style (MRS hereafter; a preference for middle response categories). Similar results linking participants' understanding or motivation with less SL were obtained by Loosveldt and Beullens (2017), Vandenplas et al (2018), andŻyczyńska-Ciołek andKołczyńska (2020). Vandenplas et al (2018) researched the interview speed and found that short (fast) interviews were associated with more SL, whereas West et al (2020) found that longer interviews were associated with a lower data quality.…”

“…In the face-to-face mode, response bias is not purely the function of respondents; it is also the function of interviewers' characteristics and behaviors (Dijkstra 1987;Klausch, Hox and Schouten 2013). Interviewers' sociodemographic traits (e.g., age and gender) and work-related traits (e.g., workload and experience), as well as the match of these traits to various respondent characteristics, can impact survey results, including the presence of RS (e.g., Beullens and Loosveldt 2016;Bittmann 2022;Życzyńska-Ciołek and Kołczyńska 2020).…”