#Papers & Studies
AI tools in academic writing: Insights from university seminars

Keywords: AI tools, academic writing, higher education, AI-assisted writing, AI integration

Exploring the integration of AI tools into academic writing, this study reveals valuable insights from seminars at the University of Bremen and Ruhr University Bochum. These seminars aimed to bridge the gap between theoretical discussions on AI and its practical applications in academic writing.

Key AI tools and their uses

Students explored a variety of AI tools, including:

  • BingAI
  • ChatDoc
  • ChatGPT (free version 3.5)
  • ChatPDF
  • Connected Papers
  • Consensus
  • DeepL Translate and Write
  • DocTranslator
  • ElevenLabs
  • Elicit
  • Humata AI
  • JenniAI
  • LightPDF
  • Litmaps
  • Open Knowledge Maps
  • Neuroflash
  • PerplexityAI
  • ResearchRabbit
  • Scribbr
  • SciSpace

These tools were employed across different writing stages: topic selection, literature search, reading, conceptual planning, text production, and revision.

Findings and observations

  • Topic selection and research: AI was minimally used for topic selection, with some students finding ChatGPT helpful for generating ideas. For content research, ChatGPT was popular but required careful prompt formulation and verification against academic sources.
  • Literature search: Tools like Connected Papers and Open Knowledge Maps were praised for visually mapping relationships between publications, though their reliability varied.
  • Literature processing: ChatPDF was useful for summarising texts but sometimes provided inaccurate or irrelevant summaries, necessitating careful review.
  • Text production: Students found limited benefit from AI-generated content for writing, preferring to use AI tools for inspiration and supplementary material.
  • Text revision: AI tools like DeepL Write were effective for stylistic improvements, though they could sometimes complicate the text unnecessarily.

Recommendations for educators

  1. Reflect on processes: Encourage students to document and reflect on their writing process and AI tool usage.
  2. Promote critical engagement: Teach students to verify AI-generated content.
  3. Balance methods: Use AI tools to complement traditional research and writing methods.
  4. Stay informed: Keep updated on AI developments and adapt teaching strategies accordingly.

Conclusion

AI tools have the potential to enhance academic writing by supporting various stages of the writing process. By fostering a balanced and critical approach, educators can help students effectively integrate AI tools into their academic work, preparing them for a future where AI is an integral part of academia.

Personal observation: The study balances leveraging AI tools for efficiency with maintaining the integrity of academic writing. It’s clear that while AI can significantly aid in various writing stages, educators must play a pivotal role in fostering critical thinking and proper usage among students.

Citation: Brommer, S. & Heimgartner, S. (2024). KI-basierte Tools beim wissenschaftlichen Schreiben.
Virtuelles Kompetenzzentrum: Künstliche Intelligenz und Wissenschaftliches Arbeiten.
DOI 10.5281/zenodo.12097856

Read the full article here (available in German).