Didaktische Studiengangentwicklung: Rahmenkonzepte und Praxisbeispiel

Salden, Peter; Fischer, Kathrin; Barnat, Miriam

doi:10.1007/978-3-658-12067-2_9

Cited by 4 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Students in particular need to be made aware of these weaknesses, as ChatGPT's quick and well-formulated answers may tempt them to overlook scientific ways of working. In addition, it must be taken into account that ChatGPT, like other language models, is based on probabilities of word sequences and therefore may output incorrect or inaccurate answers [34,35]. These study shows that ChatGPT still has weaknesses in answering questions that do not have a direct clear textbook answer, but require research and background information.…”

Section: Discussionmentioning

confidence: 92%

How the Connectedness Between Humans and Nature Can Be Increased? Comparing ChatGPT’s Language Use to Scientific Literature

Kleespies,

Schneider,

Emge

et al. 2023

Preprint

View full text Add to dashboard Cite

The artificial intelligence chatbot ChatGPT can answer questions and solve tasks posed by the user within seconds. Using an exemplary task, this study investigates the reliability and con-sistency of the answers given by ChatGPT. For this purpose, it was analyzed whether the an-swers differ when a question is asked repeatedly and whether selected content categories and associated word combinations coincide with those from scientific literature. ChatGPT was asked 400 times, how the connectedness between humans and nature can be increased. The responses were classified into a categorization system and the most frequently occurring word combina-tions for each category were extracted. These word combinations were used for a search of re-lated scientific literature. In addition, the Normalized Mutual Information (NMI) between the 400 individual responses was calculated. The results show that the phrases used by ChatGPT within the categories are also found in the scientific literature, however, the number of mentions between scientific literature and ChatGPT differs. In addition, the NMI suggested that there are large differences between the categories mentioned in the individual answers. These results in-dicate that ChatGPT currently still has difficulties in answering complex questions and cannot yet replace scientific literature research or academic writing.

show abstract

Section: Discussionmentioning

confidence: 92%

How the Connectedness Between Humans and Nature Can Be Increased? Comparing ChatGPT’s Language Use to Scientific Literature

Kleespies,

Schneider,

Emge

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Teachers cannot simply ban the use of ChatGPT and similar tools, as their use does not currently qualify as plagiarism (Salden & Leschke, 2023) and they are becoming an integral part of various industries. 4 Educators must therefore teach responsible usage of these tools, while being aware of the inherent tensions, strengths, and weaknesses in relation to research paper assignments.…”

Section: Tensions In the Use Of Chatgpt And Other Large Language Mode...mentioning

confidence: 99%

Using ChatGPT and Other Large Language Model (LLM) Applications for Academic Paper Assignments

Jungherr¹

2023

Preprint

View full text Add to dashboard Cite

Large language models (LLMs), like ChatGPT, GitHub Copilot, and Microsoft Copilot, present challenges in university education, particularly for paper assignments. These AI-driven tools enable students to (semi)automatically complete tasks that were previously considered evidence of skill acquisition, potentially affecting grading and skill development. However, the use of these tools is not legally considered plagiarism and is becoming increasingly integrated into various software solutions.University education in the social sciences aims to develop students' abilities to make sense of the world, connect their observations with abstract structures, measure phenomena of interest, systematically test expectations, and present findings in structured accounts. These practices are learned through repeated performance of tasks, such as writing research papers. LLM applications like ChatGPT create conflicting incentives for students, who might rely on them to produce parts of their papers instead of engaging in the learning process.While LLMs can be helpful tools for knowledge discovery, writing assistance, and coding assistance, using them effectively and safely requires an understanding of their underlying mechanisms, potential weaknesses, and enough domain knowledge to identify mistakes. This makes LLMs particularly challenging for students in the early stages of acquiring scientific skills and domain knowledge.Educators must enable and train students to responsibly use these new tools, reflecting on the underlying tensions and their strengths and weaknesses for academic writing tasks. This working paper aims to provide guidelines on responsible LLM use in academic contexts, specifically for students at the Chair for the Governance of Complex and Innovative Technological Systems at the University of Bamberg. The paper discusses the function of written paper assignments, the tasks necessary to complete them, and evaluates ChatGPT's performance in assisting with these tasks. It concludes with observations and advice for students to maximize the benefits of LLMs while mitigating potential risks in academic contexts, focusing on enabling learning.

show abstract

Effizienz und Effektivität von Blended-Learning-Lernverfahren

Friedrich-Haßauer

2023

Managementforschung

View full text Add to dashboard Cite

Didaktische Studiengangentwicklung: Rahmenkonzepte und Praxisbeispiel

Cited by 4 publications

References 4 publications

How the Connectedness Between Humans and Nature Can Be Increased? Comparing ChatGPT’s Language Use to Scientific Literature

How the Connectedness Between Humans and Nature Can Be Increased? Comparing ChatGPT’s Language Use to Scientific Literature

Using ChatGPT and Other Large Language Model (LLM) Applications for Academic Paper Assignments

Effizienz und Effektivität von Blended-Learning-Lernverfahren

Contact Info

Product

Resources

About