Artikel

This work proposes using the feature of large language models (LLMs) to learn about the "data bias" injected by diverse SDG classifications into bibliometric data by exploring five SDGs. We build a LLM that is fine-tuned in parallel by the diverse SDG classifications inscribed into the databases' SDG classifications. Our results show high sensitivity in model architecture, classified publications, fine-tuning process and natural language generation. The wide arbitrariness at different levels raises concerns about using LLM in research practice.

In the Dutch academic system, PhD theses can be awarded with cum laude and at most 5% of all PhD graduates receive this selective distinction for their thesis. In this paper, we investigate whether there is gender bias in awarding cum laude, using data from one of the major Dutch research universities. A main result is that the set of PhD theses receiving cum laude on average do not have a higher quality than the best theses not getting cum laude. A second main result is that, after controlling for the quality of the PhD theses, women still have a substantially lower probability to receive cum laude. These results strongly suggest that the distribution of awards suffers from gender bias.

Based on a multimodal understanding of governance in German higher education, thsi paper investigates the multiple objectives and factors that shape the digitalization of higher education internationalization. We find that while values-based and competition-related approaches often motivate higher education institutions to embrace digitalization, the implementation is rather shaped by hierarchy and collegial governance factors. We suggest understanding digitalization as a comprehensive approach intertwined with all levels and modes of governance in higher education.