Module 5: Ethical aspects of AI-aided content analysis


Lesson 5.1: Key takeaways and ethical perspectives

AI-aided content analysis of sustainability communication

nils.holmberg@iko.lu.se

Key takeaways and ethical perspectives

  • Understand how AI can support content analysis in sustainability research
  • Gain practical skills in text and image data analysis using low-code tools
  • Learn to critically interpret results produced by machine learning models
  • Explore ethical boundaries of automation in communication analysis
  • Reflect on transparency, bias, and accountability in AI-supported research

What You Learned in This Course

  • Collect and organize sustainability content from real-world sources
  • Analyze both textual and visual communication using Python notebooks
  • Apply NLP and computer vision techniques in a social science context
  • Summarize and visualize data to uncover communication patterns
  • Connect computational methods to audience effects and strategic messaging

Ethics of Automating Content Interpretation

  • Question how algorithmic bias might shape findings
  • Evaluate the limits of interpretability in automated results
  • Discuss the role of the researcher in guiding and validating AI output
  • Consider privacy and consent when collecting public digital content
  • Integrate ethical reflection into research design and dissemination

Analysis documentation and open science (Jupyter)

  • Emphasize reproducibility through transparent workflows
  • Store data, code, and results in accessible formats
  • Use cloud platforms and version control for collaborative research
  • Frame analysis in ways that support peer review and replication
  • Connect computational practice with open science values
  • computer lab: download ipython notebook from google colab

Using Jupyter Notebooks to Ensure Reproducibility

  • Quarto integrates code, output, and documentation in one file
  • Supports reproducible research with automated execution
  • Allows easy conversion to HTML, PDF, or Word for publication
  • Embeds metadata, citations, and environment settings
  • Ideal for sharing transparent, well-structured analytical reports

Communicating AI-aided content analysis (Quarto)

  • Translate technical findings into accessible narratives
  • Use visualizations to highlight patterns in sustainability messaging
  • Connect communication effects with organizational strategies
  • Present methods and results clearly for both technical and non-technical audiences
  • Position findings within broader ethical and scientific discussions
  • computer lab: convert ipython notebook to quarto document

Turning Your Analysis into a Research Publication

  • Start with a well-documented Google Colab or Quarto notebook
  • Define a clear research question with reproducible methods
  • Use structured visualizations to support interpretations
  • Align findings with existing literature and frameworks
  • Submit to journals focusing on digital methods, media studies, or sustainability communication