Jour Fixe 131 | Ságvári Bence, Stefkovics Ádám: How do people answer web surveys? Distractions, multitasking, and the context of completion and their consequences in a web survey

   7th March 2024 - 7th March 2024

The HUN-REN Centre for Social Sciences Institute for Sociology cordially invites you to

How do people answer web surveys? Distractions, multitasking, and the context of completion and their consequences in a web survey

 

Lecturers: Bence Ságvári (HUN-REN CSS RECENS); Stefkovics Ádám (HUN-REN CSS RECENS)

Discussants:  Ferenc Mújdricza (KSH); Gábor Hajdu (HUN-REN CSS Institute for Sociology)

Date: 07 March 2024, 1 pm

Venue: HUN-REN Centre for Social Sciences Institute for Sociology, 1097 Budapest, Tóth Kálmán utca 4. Room B1.15 

Online: Zoom link: 

https://us06web.zoom.us/j/88920559667?pwd=DVJefnqzw27tKMNq2SQXUt9SUx7l6N.1

Meeting ID: 889 2055 9667
Passcode: 502566

 

Abstract:

As opposed to interviewer-administered surveys, researchers have limited control over the response process in web surveys. Participants in online surveys can answer the questions from any location, at any time, using different devices, and often engage in other activities while completing the survey. This variability in response conditions can affect the quality of the data collected. This study contributes to this problem in two ways: first, by developing a theoretical framework that shows how different aspects of survey completion affect the four stages of the response process, and second, by providing empirical evidence from a comprehensive set of completion-related questions collected in an online survey in Hungary. Using cluster analysis, we identify five typical groups of respondents. The results show that many respondents participated in the survey under conditions that could be considered suboptimal, such as using a small device, during the night, in places with little privacy, or when heavily multitasking. Nevertheless, we found little evidence that these circumstances were related to data quality (i.e. duration, item non-response, straightlining, inconsistent responses), but only to reported noise and concentration levels. Despite this, we advise researchers to routinely monitor completion behaviour in web surveys.