26.2. 2019, 13:00 Uhr
Mannheim, B2,8 (links)
The increasing use of smartphones opens up opportunities for novel ways of survey data collection, but also poses new challenges for researchers. We can collect location tracking data and ask intensive longitudinal questions, but the questions we ask can get increasingly intrusive and we risk over-asking participants. In order to investigate the effects of taking a traditional diary for the Dutch Time Use Survey to a mobile setting, we studied nonresponse and nonresponse bias in survey questions, smartphone time use diary data and recording sensor data (GPS locations and call data). Results show that the correlates of nonresponse are somewhat different for every task, and that nonresponse (bias) affects survey estimates as only a specific group is willing to participate in the smartphone parts.
In addition, smartphones thus enable the passive collection of sensor data (for example location tracking data) alongside to survey participation. Linking GPS data to self-reported time use surveys can be valuable for understanding how people spend their time, as location data add context to people’s reports about their time use. We investigated whether and how the passive collection of geographical location data (coordinates) proves useful for deriving functional locations of respondents. We recorded at what location every activity took place, and matched this to GPS locations. Results show that a lot of measurement errors occur in the location tracking data making it difficult to record locations.
In this talk, I present the results of these two research projects on smartphone-only time use research. The talk will focus on what we have learned, what research remains to be done, and what the potential is for smartphone-only time use research.
About the Speaker
Anne Elevelt is a PhD Candidate in the field of survey methodology at the Department of Methodology and Statistics at Utrecht University. Her research project focuses on smartphone only-survey research, the use of sensor data, and the impact these two have on nonresponse (bias) and data quality.