Thijs Carrière

Department of Methodology and Statistics
Faculty of Behavioral and Social Sciences
Utrecht University
Email
Website

Project
Data quality in app based data collection

Over the past decades, the field of survey methodology has changed rapidly as result of technological advancement. New data collection methods (e.g., web surveys, mobile apps, sensors, data donation) are more and more commonly used to collect large amounts of new, granular data, which can be used to augment survey data. However, these new data types and new collection methods bring new methodological challenges to data collection and data quality. One such challenge is keeping participants involved with these data collection methods, as they might be burdensome or participants might have privacy concerns. Selective participation in these novel data collection methods can lead to biased data and incorrect inferences.

This PhD project will focus on mobile apps for data collection. In the project, we will investigate what app features can contribute to an increase in study participation and data quality, such as (1) more agency for participants over when data is collected, (2) provision of personalized feedback, and (3) the ability to alter collected data. The studies and findings are aimed to guide the development of a participant centered app as part of the VIDI project of Bella Struminskaya.

The goal of the PhD is to understand how participants can be engaged and committed to app based data collection, and how app design choices compromise with the quality of the obtained data.

In the first project, I will investigate the effects of personalized feedback on participant behavior in app-based data collection. App paradata will be used to get a more nuanced view on participant behavior (e.g., is provided feedback actually viewed by participants). Both potential positive effects (higher study adherence) and negative effects (unwanted behavioral change) are investigated. The goal is to phrase concrete recommendations on when and how to provide personalized feedback in data collection.

In the second project, a literature study on existing research apps will be conducted. The aim is to get an overview on what kind of app functionalities are currently used in research apps and the effects these are supposed to have on study participants. Together with the first project, this project should lead to concrete recommendations on what app features to include when building the Vidi app.

The third project will continue on a first version of the Vidi research app. We will study the willingness to participate in app based data collection in an existing research panel. To get an more realistic scenario, a first app version will be used for this, or alternatively, a vignette study will be conducted. Results will be used to further advance the research app functionalities.

The fourth project will look into the data quality of data collected in app based data collection with more participant agency over the data (e.g., being able to stop collection temporarily, or being able to edit/remove collected data). Ideally, this study will make use of an applied data collection.

Supervisors
Dr. Bella Struminskaya
Dr. Laura Boeschoten
Dr. Peter Lugtig

Financed by
NWO VIDI

Period
January 2024 – December 2027