My research focuses on quantitative social research methodology, and I use machine learning to identify bias in social science data, especially due to measurement and nonresponse error, and to process unstructured data. I also use LLMs for the classification of text data, and explore strategies to use AI in questionnaire design.
I first started to study engineering, but while I enjoyed the math component of the program, my interest was in learning about how our society works. Thus, I changed the field to social sciences and humanities, while still keeping my enthusiasm for statistics. As a quantitative social scientist I can now follow both of my passions at the same time, statistics and social research.
Data science and AI offer new amazing possibilities for social science research to discover, analyze and combine otherwise not available data and improve social science methodology. They also offer new opportunities of interdisciplinary collaboration (e.g., between social science and computer science), which are more likely to result in break-through discoveries.
In two projects with colleagues from Germany and Canada, we used LLMs to classify open-ended answers in surveys to better understand reasons why people are willing or not willing to share additional data (e.g., web-paradata, social media data, health data data, geo-location data) within a survey. We discovered interesting patterns regarding privacy and trust, which will use to inform future data collections.
Something interesting about myself- I started to play chess at the age of six, and I’m still a passionate player. Playing international tournaments allowed me to make friends around the world.