52442 research outputs found
Sort by
Holy days, lost days?
Get PDFDo public holidays meaningfully affect economic output? In Germany, strict Sunday laws create a unique natural experiment: when public holidays fall on Sundays, they typically do not additionally disrupt business activity. Exploiting this variation across states and years, I estimate the economic cost of a “lost” workday. Using monthly manufacturing data and a stacked event-study approach, I find that weekday holidays lead to modest but measurable reductions in output. Scaling the estimates implies annual GDP losses between 0.06% and 0.28%, depending on whether the effect is assumed to apply only to manufacturing or to the whole economy
United in diversity? Contextual biases in LLM-based predictions of the 2024 European Parliament elections
No full text
News without borders: Domain adaptation of multilingual sentence embeddings for cross-lingual news recommendation
No full textHow do AI educators use open educational resources? A cross-sectoral case study on OER for AI education
Get PDFArtificial Intelligence (AI) literacy is essential for society as a whole. While general frameworks and resources to support self-directed learning on AI are widely available, research on how to support AI educators, particularly those without AI expertise (non-experts), using external materials and resources is relatively scarce. This article explores the potential of open educational resources (OER) to enhance AI education, with a specific focus on the requirements and practices of AI educators. Through a case study of the AI Campus learning platform, the article examines how educators from diverse sectors such as school education, higher education and professional education utilise OER for AI education. The study aimed to identify patterns of OER usage, AI educator motivations and the sector-specific integration of OER into teaching practices. A survey study of 260 educators from Germany, Austria, and Switzerland using AI Campus content revealed that educators prefer smaller, modular OER formats and value suitable, high-quality and accessible content. The reputation of the person or institution that created the OER content does not seem to play a major role. Sector-specific differences could be observed in particular with regard to full online courses, face-to-face learning scenarios and the AI learning objectives of an educator. By focusing on educators’ perspectives, the study provides insight into how AI education can be strengthened across sectors through the use of OER materials and ultimately benefit learners through suitable, high-quality content and adequate AI learning scenarios
