SODAS Data Discussion 1 (Fall 2025)
Copenhagen Center for Social Data Science (SODAS) aspirers to be a resource for all students and researchers at the Faculty of Social Sciences. We therefore invite researchers across the faculty to present ongoing research projects, project applications or just a loose idea that relates to the subject of social data science.
Two researchers will present their work. The rules are simple: Short research presentations of ten minutes are followed by twenty minutes of debate. No papers will be circulated beforehand, and the presentations cannot be longer than five slides.
Discussion 1
Presenter: Berit Tricia Heling
Title: Exploring the Psychometrics of Employer Image Attributes with LLMs
Abstract:
In response to growing talent shortages, organizations try to improve their Employer Branding strategies in job advertisements to attract qualified candidates. Understanding how the Employer Image is communicated and the outcomes associated with its presence and valence remains a complex and resource-intensive task, often constrained by the volume of job ads and the limitations of human raters. Natural Language Processing (NLP) through Large Language Models (LLMs) has emerged as a tool with significant potential in the analysis of large amounts of text. Drawing on extensive datasets of job advertisements and corresponding recruitment outcomes, we investigate the reliability and validity of LLM-generated ratings in the domain of Employer Branding. Preliminary findings raise questions about the correspondence of human and LLM perception, training methodologies of LLMs, and implications for the usage of LLMs to analyze Employer Image attributes in job advertisements.
Discussion 2
Presenter: August Lohse
Title: What Can We Learn from 10 Years of Student Exams?
Abstract:
In this presentation we will discuss a unique dataset of student handins. I will present two use cases of the data - specifically on how the emergence of widely available GAI has affected the writing of students, as well as an analysis of how writing complexity and grading are related. Afterwards there will be a lot of time for discussing new use cases of the data, and develop research ideas.