276557 research outputs found
Sort by
Towards reliable generative AI-driven scaffolding: Reducing hallucinations and enhancing quality in self-regulated learning support
No full textContains fulltext :
324337.pdf (Publisher’s version ) (Open Access)Generative Artificial Intelligence (GenAI) holds a potential to advance existing educational technologies with capabilities to automatically generate personalised scaffolds that support students' self-regulated learning (SRL). While advancements in large language models (LLMs) promise improvements in the adaptability and quality of educational technologies for SRL, there remain concerns about the hallucinations in content generated by LLMs, which can compromise both the learning experience and ethical standards. To address these challenges, we proposed GenAI-enabled approaches for evaluating personalised SRL scaffolds before they are presented to students, aiming for reducing hallucinations and improving overall quality of LLM-generated personalised scaffolds. Specifically, two approaches are investigated. The first approach involved developing a multi-agent system approach for reliability evaluation to assess the extent to which LLM-generated scaffolds accurately target relevant SRL processes. The second approach utilised the "LLM-as-a-Judge" technique for quality evaluation that evaluates LLM-generated scaffolds for their helpfulness in supporting students. We constructed evaluation datasets, and compared our results with single-agent LLM systems and machine learning approach baselines. Our findings indicate that the reliability evaluation approach is highly effective and outperforms the baselines, showing almost perfect alignment with human experts' evaluations. Moreover, both proposed evaluation approaches can be harnessed to effectively reduce hallucinations. Additionally, we identified and discussed bias limitations of the "LLM-as-a-Judge" technique in evaluating LLM-generated scaffolds. We suggest incorporating these approaches into GenAI-powered personalised SRL scaffolding systems to mitigate hallucination issues and improve the overall scaffolding quality.19 p
Home literacy and numeracy precursors of Chinese kindergarteners' early literacy and mathematics achievements
No full textContains fulltext :
325753.pdf (Publisher’s version ) (Closed access)This study examined the stability and change in the home learning environment and its longitudinal relationships with children's literacy and mathematics achievements in mainland China. In addition, we explored the cross-domain relations of the home literacy environment and early mathematics. One hundred and twenty-nine kindergarten children were assessed on early literacy (syllable/tone awareness and Pinyin knowledge) and mathematics skills in kindergarten 2 (K2) and 3 (K3), and their parents completed a questionnaire on the home learning environment in K2 and K3. Results showed a significant correlation between the home learning environment from K2 and K3, indicating stability. There was an increase in Chinese writing/speaking activities, digital home literacy activities, traditional home numeracy activities, and parental expectations. In contrast, there were no increases in informal home literacy activities and digital home numeracy activities. Chinese writing activities in K2 and K3 predicted children's Pinyin knowledge in K3. Parental numeracy expectations predicted early mathematics skills in K3. Finally, a cross-predictive role of Chinese writing activities in basic arithmetic skills was shown. These findings contribute to the research by indicating that home activities focusing on the codes of written language are more beneficial for literacy development than activities centered on speaking for Chinese children. Exposure to writing experience could also enable a promotion in early mathematics. Furthermore, parental numeracy expectations were a unique predictor of early mathematics skills. Our findings provide support that the home learning environment is related to children’s early academic skills and suggest that this relationship is shaped by cultural context.13 p
Exploring proof autoformalization with Mistral on Herald
No full textItem does not contain fulltextCICM 202
From Mosquitoes to Birds: Unravelling the Role of Avian Hosts in Vector-Borne Pathogen Dynamics
No full textContains fulltext :
326241.pdf (Publisher’s version ) (Open Access)"This thesis examines how vector-borne pathogens (VBPs) are emerging and spreading in increasingly urbanized environments, and how these changes affect wild bird populations. Early detection and surveillance are shown to be essential for managing risks to wildlife, domestic animals, and humans.
Using public reports of dead birds, chapter 2 demonstrates that citizen-science data can effectively track pathogen emergence. The mosquito-borne virus USUV spread rapidly across the Netherlands, moving from the southeast to the north within three years and causing a 30% decline in Blackbird populations. Such data proved valuable for modelling transmission dynamics and linking pathogen pressure to wildlife population declines.
Chapter 3 extends these findings by showing that other resident or short-distance migratory species, Song Thrush, Dunnock, Eurasian Wren, and Eurasian Robin, also experienced population declines associated with USUV. Long-distance migrants were unaffected, suggesting they play a limited role in year-round VBP transmission.
Chapters 4 and 5 explore why resident birds may be key contributors to pathogen maintenance. Movement analyses revealed substantial spatial mixing: juveniles dispersed widely, while adults frequently moved between urban and non-urban habitats, likely tracking seasonal resources. These movements can expose urban populations to new pathogens and shape local transmission patterns.
Finally, the thesis investigates how urban environments affect host immunity. Adult Blackbirds showed altered immune profiles in urban areas, potentially reflecting upregulated defences in response to urban stressors. No such patterns were found in nestlings, and no trade-offs between body condition and immunity were detected.
Overall, this work highlights how urbanization, bird movement, and immune responses interact to influence VBP transmission. It underscores the importance of continued surveillance and offers insights for both public health and avian conservation."Radboud University, 15 januari 2026Promotor : Foppen, R.P.B. Co-promotores : Matson, K.D., Jeugd, H.P. van der232 p
The impact of dyads and extended networks on political talk: A factorial survey experiment in the Netherlands
No full textContains fulltext :
326452.pdf (Publisher’s version ) (Open Access)Open political dialogue is regarded as foundational to democratic health and our social fabric. Here, we study political dialogue by examining with whom we prefer to talk about politics and why. In so doing, we develop and test hypotheses about what random encounters (e.g., meeting similar versus dissimilar others, meeting friends or colleagues) foster political dialogue, pose and test conjectures about what features of extended networks facilitate political debate, and present a new unique factorial survey experiment to answer our questions. We incorporated this factorial survey experiment within the NEtherlands Longitudinal Lifecourse Study 2022 (NELLS) and presented to a large sample of Dutch citizens - including Dutch ethnic majority members as well as minoritized Dutch with a Turkish or Moroccan heritage - a choice to engage in political talk or not. Hierarchical linear models reveal that relationship strength, rather than identity similarity (e.g., gender, ethnic), is the primary driver of opting for political dialogue. However, in encounters lacking prior relationship history, gender similarity increases the willingness to engage, and similar political views do relate to engaging in substantive debate modeled dichotomously. Ethnic diversity within extended networks fosters political discussion, while network size has a nonlinear association - medium-sized networks are more conducive to dialogue than very small or very large ones. These findings contribute to debates on political polarization by highlighting the relational conditions that encourage political exchange.14 p
Machine learning for quantifier selection in cvc5
No full textContains fulltext :
326608.pdf (Publisher’s version ) (Open Access
Functional characterization of pathogenic SATB2 missense variants identifies distinct effects on chromatin binding and transcriptional activity
No full textContains fulltext :
326079.pdf (Publisher’s version ) (Open Access)11 p
Methods for investigating electronic health records and patient-provider interaction: A scoping review.
No full textContains fulltext :
324823.pdf (Publisher’s version ) (Closed access)OBJECTIVE: Electronic Health Records (EHRs) have improved many aspects of healthcare but also introduced challenges, including increased documentation and reduced patient-provider interaction. With recent artificial intelligence (AI) developments in the EHR landscape, such as Automated Medical Reporting (AMR) technologies that aim to automate visit summarization, there is a need to understand how these innovations affect medical interactions. However, methodological inconsistencies in existing studies on EHR use and medical interactions limit study comparability. This scoping review identifies methods to study the impact of EHR use on patient-provider interaction, providing a foundation for future evaluations of AMR implementations. METHODS: A systematic search identified empirical studies (2015-present) on patient-provider interaction involving EHRs, supplemented by three prior reviews. Fifty-nine studies met the inclusion criteria. Data extraction focused on research methods, concepts, indicators, and measurement tools used. RESULTS: The review identified diverse methods to study EHR use within medical interactions, including video/audio analysis, structured observation, surveys, and interviews. Studies focused on concepts such as patient-centered communication, attention, satisfaction, computer interaction, and room layout, using diverse standardized and self-developed instruments. Most studies treated EHR use as separate from medical interaction. A smaller group used qualitative methods, such as conversation analysis and ethnography, to explore how EHR use is embedded in medical interaction. Indicators included gaze, typing, talk time, and consultation length. CONCLUSION: Two dominant approaches emerged: one treating EHR use as separate from medical interaction, and another, less common, viewing both as functionally integrated. While various concepts and instruments were identified, many studies relied on unvalidated, non-experimental methods, limiting comparability and causal inference. A key gap remains in standardized methods that link broad communication concepts with specific functional behaviors. As AI-driven tools like AMR emerge, future research needs to adopt more integrative approaches to assess their impact and support human-centered EHR design.01 januari 202
Retirement and life course transitions: Inequality and the welfare state in Europe
No full textContains fulltext :
324931.pdf (Publisher’s version ) (Closed access)Radboud University, 28 januari 2026Promotor : Kraaykamp, G.L.M. Co-promotor : Visser, M.193 p
The Journal Attention Cycle: Indicators as Assets in the Chinese Scientific Publishing Economy
No full textContains fulltext :
326520.pdf (Publisher’s version ) (Open Access