Forschungsinformationssystem der Universität Bamberg
Not a member yet
66317 research outputs found
Sort by
Exploring the Hypothetical Impact of Genetic Engineering on Ethnicity : An Analysis of a Large‐Scale Data Set Retrieved From a Museal Setting
Full text linkCritics of human genetic engineering warn that if ever put into practice, this will diminish human diversity, especially regarding skin color. Nonetheless, given the solid and shameful causal link between skin color and discrimination, the provocative question is whether to manipulate this feature and create children whose stereotype-aligning features reduce the risk of evoking hostility in the social environment. To address this possibility, we analyzed data from an interactive exhibit in a German museum that partly addresses these questions. Visitors could manipulate randomized features of a virtual child—for example, appearance and intelligence—to align them with their notion of a “perfect child.” Analysis of N = 13,641 virtual children showed an apparent effect on aligning skin color with a Caucasian type. This was true for extreme light and dark, randomly assigned initial skin colors, but stronger for the latter. This preference could reflect the attempt to align the hypothetical child's skin color with the creating visitors. We also analyzed the chosen skin-color-dependent distribution of designed intelligence based on previous findings showing that high intelligence is less desirable for Black than White persons. We revealed that virtual children with a chosen darker skin color were designed with relatively lower intelligence and a larger proportion of maximized and minimized values. Although most effects were small, they might indicate racial prejudices and/or the attempt to design virtual children with high alignment with normative stereotypes. Our findings provide an important starting point to empirically inform the critical and timely debate about human genetic engineering
Knowing when it’s ‘necessary’ : academic parents’ strategic intervention and involvement
Full text linkThis paper investigates how parents from different socioeconomic backgrounds support their teenage children in upper secondary education and how this influences their educational pathway. Based on 31 in-depth qualitative interviews with former participants of the BiKS-panel-study, this study combines qualitative insights with longitudinal data to examine mechanisms of educational inequality. While much research contrasts broader social groups, this paper also uncovers in-group-variation. The results reveal differences not only between non-academic and academic parents but also within the group of academic parents, especially between educational insiders (teachers) and outsiders (parents with university degrees in other subjects). Educational insider parents can better recognize educationally relevant opportunities, situations, and decisions and provide appropriate support because they have access to more relevant resources and field-specific knowledge. They do not intervene consistently but instead recognize when intervention is ‘necessary’ within the specific educational system, resulting in educational advantages for their children
Connecting the dots : how social networks shape macroeconomic expectations
Full text linkStandard behavioral frameworks often overlook how social network topology shapes the coordination of inflation expectations. To address this, I build a hybrid agent-based model that integrates “narrative-rooted” heuristics within a standard New Keynesian structural scaffold. In this framework, beliefs evolve through dual channels: performance-based selection (heuristic switching) and social diffusion (DeGroot learning). Simulations across canonical topologies reveal that seeding a target-based narrative in high-centrality nodes compresses forecast dispersion and accelerates convergence. However, a structural trade-off emerges: while performance evaluations dampen distorting narratives, strong social persuasion reduces volatility but simultaneously decouples expectations from economic fundamentals. This occurs because agents prioritize peer alignment over objective data, highlighting the double-edged nature of social networks: they stabilize expectations when credible narratives diffuse through hubs, but sever the link between beliefs and fundamentals when conformity overrides economic signals
The Influence of Country Regulations on the Relationship between the Child-Staff Ratio and Process Quality in Childcare Centers
Full text linkDer Personalschlüssel ist eines der meistuntersuchtesten Strukturqualitätsmerkmale zum Einfluss auf die Prozessqualität in Kitas. Internationale Forschungsbefunde legen nahe, dass die Prozessqualität in Kita-Gruppen von bestimmten Grenzwerten des Personalschlüssels abhängt. In vielen Ländern existieren daher auf politischer Steuerungsebene Vorgaben zur maximal zulässigen Anzahl der Kinder pro pädagogischer Fachkraft in Form von allgemeinen Empfehlungen oder offiziellen Richtwerten. Die vorliegende Metaanalyse zielt darauf ab, die Bedeutung solcher Ländervorgaben für den Zusammenhang des Personalschlüssels und der Prozessqualität in Kitas anhand von Moderatorenanalysen zu untersuchen. Die Ergebnisse der 139 Effektstärken aus 45 Publikationen und 14 Ländern belegen, dass ungünstigere Personalschlüssel mit einer geringeren Prozessqualität in altersgemischten Gruppen und Gruppen für unter Dreijährige zusammenhängen. Die Existenz von Landesvorgaben zum Personalschlüssel, die wissenschaftlichen Empfehlungen entsprechen, weist nur in altersgemischten Gruppen auf geringere Zusammenhänge hin. Die Ergebnisse werden mit Blick auf die Definition von Ländervorgaben und deren Relevanz, besonders für altersgemischte Gruppen und jüngere Kinder, diskutiert.Child-staff ratios are one of the most studied structural quality characteristics in terms of their influence on process quality in childcare centers. International research findings suggest that the process quality in childcare groups depends on specific threshold values of the child-staff ratio. In many countries, official guidelines exist that impose restrictions on the maximum number of children per staff member in the form of general recommendations or official regulations. This meta-analysis examines the importance of such regulations and whether they moderate the relationship between child-staff ratios and process quality in childcare centers. The results of the moderator analysis across 139 effect sizes from 45 publications and 14 countries show negative correlations between less favorable child-staff ratios and process quality in mixed-age groups and groups for children under 3 years. However, the existence of country regulations for child-staff ratios corresponding to scientific recommendations only points to lower correlations in mixed-age groups. Based on these results, the authors discuss the definition of country regulations and their relevance, particularly for mixed-age groups and younger children
Entwicklung und Validierung eines Instruments zur Erfassung selbstregulierten Lernens in Lernprodukten
Full text linkDie Analyse von studentischen Lernprodukten ermöglicht eine verhaltensnahe Diagnostik von kognitiven Lernstrategien und liefert eine Grundlage für die Evaluation von Lernaktivitäten. In der vorliegenden Studie wird untersucht, ob Organisations- und Elaborationsstrategien anhand von Lernprodukten reliabel erfasst und am Lernerfolg validiert werden können. Mit einem neu entwickelten Codierschema wurden 229 Vorlesungszusammenfassungen von 24 Lehramtsstudierenden anhand verschiedener hoch-inferenter Kategorien analysiert, die eine effiziente Auswertung ermöglichen. Für tiefere Einblicke wurden darüber hinaus 94 Vorlesungszusammenfassungen von 10 Studierenden mit besonders hohem bzw. niedrigem Lernerfolg mit niedrig-inferenten Kategorien ausgewertet. Die Ergebnisse zeigen für die hoch-inferenten Kategorien eine substanzielle und für die niedrig-inferenten Kategorien eine sehr gute Interrater-Reliabilität. Es konnten signifikante Korrelationen zwischen der Nutzung von Lernstrategien im Lernprodukt und dem Lernerfolg (Prüfungsergebnis) der Studierenden nachgewiesen werden. Studierende mit hohem Lernerfolg nutzten Lernstrategien fast aller Kategorien häufiger als solche mit niedrigem Lernerfolg. Grenzen und Potentiale des Ansatzes für die Diagnostik und Evaluation selbstregulierter Lernprozesse werden diskutiert.The analysis of students’ learning products enables a behavioral diagnosis of cognitive learning strategies and provides new information for the evaluation of learning activities. The current study aims to analyze whether organization and elaboration strategies can be measured reliably in learning products and can be validated by learning success. With a newly developed coding scheme 229 lecture summaries of 24 student teachers were examined by different high-inferent categories, which enable an efficient evaluation of the learning products. For deeper insights, 94 lecture summaries of 10 students with either high or low learning success were analytically rated by means of low-inferent categories. Overall results show an interrater reliability that is substantial for high-inferent categories and very good for low-inferent categories. Significant correlations between the usage of learning strategies for the learning products and learning success (examgrades) could be found. Students with high learning success used learning strategies of nearly all categories more frequently as those with low learning success. Limitations and potentials of the approach for diagnosing and evaluating self-regulated learning processes are discussed
Impulse zu Informationskompetenz im Zeitalter von KI : ECIL vom 22. bis 25. September 2025 in Bamberg
Full text linkVom 22. - 25.09.2025 hat die Universitätsbibliothek Bamberg die 9. European Conference on Information Literacy (ECIL) ausgerichtet. Das Hauptthema 2025 war “Information Literacy in an AI-driven World”, und Diskussionen über KI und dessen Auswirkungen waren in vielen Beiträgen ein Kernthema. Dennoch wurden im viertägigen Programm zahlreiche Aspekte und Perspektiven zur Informationskompetenz beleuchtet. Neben konzeptionellen Beiträgen und der Präsentation von Befunden empirischer Studien bot das Programm Workshops, in denen Teilnehmende bspw. Tools zur Informationskompetenz-Vermittlung testen durften, oder neue pädagogische Vermittlungsansätze detaillierter vorgestellt und diskutiert wurden. Insgesamt gab es drei Keynotes, zwei eingeladene Gastvorträge, 88 Einzelbeiträge, 33 Best Practice-Ansätze, 8 Kurzvorträge alá Pecha Kucha, 18 Posterpräsentationen, 14 Workshops und 4 Panels.
Mit ihren vielfältigen und internationalen Perspektiven und Diskussionen hat die ECIL 2025 in Bamberg wichtige Impulse für die Weiterentwicklung der Informationskompetenz im Kontext von KI gegeben – und gezeigt, dass Informationskompetenz auch in Zukunft ein zentrales Bildungsanliegen bleibt
Inter-organizational collaborations in open-source software ecosystems
Full text linkContext:
Open-source software (OSS) ecosystems are pivotal to modern software development by fostering innovation through inter-organizational collaboration. While becoming increasingly important, research on the patterns and dynamics of this collaboration in OSS remains limited.
Objective:
This study investigates inter-organizational collaboration within different OSS ecosystems, focusing on identifying key contributing and influential organizations, analyzing collaboration patterns, and understanding their evolution over time.
Method:
An exploratory case study with a mixed-methods approach was conducted, with data collected from five prominent OSS ecosystems on GitHub (React, Vue, TensorFlow, Bootstrap, and Flutter) involving 9947 developers and 339 organizations over ten years. Both qualitative and quantitative analyses were performed, including software repository mining and applying different levels (micro/macro) and dimensions (rational/structural/functional) of social network analysis (SNA) to examine collaboration patterns.
Results:
Key organizations, such as Alphabet, TensorFlow, Meta, Nvidia, Intel, MobileIron, IBM, and AMD, emerged as significant contributors, acting as central hubs for collaboration within OSS ecosystems. The collaboration networks showed an initial phase of rapid growth followed by stabilization, indicating project maturation. Notably, competing firms were found to contribute to the same ecosystems, underpinning the inherently cooperative, reciprocal, and multiplex nature of OSS development.
Conclusion:
The findings highlight the pivotal role of top-contributing organizations in fostering collaboration and driving innovation within OSS ecosystems. Understanding these dynamics provides valuable insights for developing strategies to enhance the effectiveness, scalability, and long-term sustainability of OSS ecosystems, benefiting both individual contributors and participating organizations
Positionspapier : Fachspezifische Informationskompetenz-Vermittlung und Künstliche Intelligenz
Full text linkDas Positionspapier erörtert Perspektiven für die Vermittlung von Informationskompetenz im Kontext von Künstlicher Intelligenz
Organizational Influence on Security Development in Open-Source Software Projects
Full text linkIncreasing technological complexity, intensified competition, and security requirements have driven open-source software (OSS) projects to become a crucial part of organizations' software development. This study focuses on the OSS project TensorFlow (TF) and uses a case study to examine how organizations and their associated developers collaborate to identify, fix and prevent security vulnerabilities. Social Network Analysis (SNA) of archived security data from software repositories is used to gain insight into security activities. The study examines the internal structure and evolution of security code collaboration, organizational networks, and top organizational contributors to TF. It also examines productivity, homophily, development diversity, and turnover rates among developers across various software releases. The in-depth insights from this research enhance our understanding of collaborative patterns in OSS communities within open software ecosystems, particularly in the security context
Determinants and effects of green bond issuance : Environmental awareness, ecological budget, biodiversity, oil and lithium
Full text linkThis study explores the determinants of green bond issuance and its environmental impacts. The empirical analysis employs a panel dataset from 29 OECD countries for the period 2014–2020 and two main models. The first model identifies determinants of green bond issuance, revealing that higher environmental awareness, GDP per capita, oil prices, a higher degree of urbanization, and lower lithium prices are associated with a higher volume of issued green bonds. The second model employs several indicators of environmental performance as dependent variables. The main independent variables are the issued green bond volume, environmental awareness, GDP per capita, environmental policy stringency, renewable energy capacity, and the share of protected areas. The results show that green bond volume, stringent environmental policy, and higher environmental awareness are positively related to the ecological budget and biodiversity while reducing the ecological footprint. Channels for this impact are positive relationships between green bond funding and renewable energy capacity and the share of protected areas. Governments should therefore not only promote the issuance of green bonds, which are essential to raise the financial resources needed to finance environmentally friendly projects, but also offer tax and investment incentives, provide technical assistance, and simplify procedures for project implementation. In addition, resources should be devoted to raising environmental awareness among the population