Integrating Generative AI into the Research Cycle: Enhancing Discovery, Analysis and Dissemination
Format: 1-2 days, online or on-site
Generative AI is reshaping how research is conducted—from idea generation to publication. This interactive workshop explores how AI-based tools can enhance each stage of the research cycle, improving efficiency, creativity, and decision-making.
In this workshop, you will engage with cutting-edge AI applications*, including Open AI’s ChatGPT, Anthropic’s Claude, Perplexity.ai, and GitHub Copilot to streamline your research processes. Through a mix of hands-on exercises, demonstrations, and discussions, this format will equip you with practical strategies for integrating AI into your work routines while maintaining good scientific practice.
The specific learning goals and topics of the workshop are:
- From Observation to Hypothesis: Use AI to brainstorm, explore research questions and refine ideas.
- Literature Review & Knowledge Synthesis: Leverage AI-driven search engines and research tools for deeper, more efficient literature exploration.
- Proposal & Manuscript Writing: Employ AI as a writing assistant for structuring proposals, refining arguments, and improving clarity.
- AI for Data Analysis & Programming: Get started with or enhance data analysis and programming skills with AI-based tools – for qualitative as well as quantitative research.
- Publishing & Science Communication: Utilize AI to craft compelling narratives for peer-reviewed publications and public engagement.
- Workflow Automation: Use agentic AI to outsource multi-step workflows while maintaining scientific rigor through quality control and verification.
- Ethical Considerations & Responsible AI Use: Discuss the challenges of AI in research, including bias, hallucinations, data privacy, and copyright within the framework of Good Scientific Practice.
*We will select specific tools just before the workshop to ensure the most current developments in this fast-evolving field are incorporated.
Generative AI Tools for Quantitative Data Analysis and Programming with Python
Format: 1.5-2 days, online or on-site
This workshop teaches you how to leverage generative AI tools to analyze data with Python—whether you’re a complete beginner, currently use other statistical software, or are an experienced Python programmer looking to enhance your workflow.
What You Will Learn
- Write Python Code Without Prior Experience—or Advance to Expert Level: Use AI assistants and copilots (e.g., GitHub Copilot, Claude, ChatGPT) to write and debug Python scripts by describing your analysis needs in natural language. Learn how to translate research questions into functional code.
- Understand, Debug, and Improve Your Code: Use AI to explain existing scripts, add documentation, troubleshoot errors, and refactor code for clarity and efficiency.
- Complete Data Analysis Workflows with AI Support: Master the entire data analysis pipeline: data cleaning, exploratory analysis, statistical testing, modeling, and visualization. Create publication-ready figures and reproducible analysis workflows.
- Navigate Limitations and Pitfalls: Understand when AI coding assistants excel and when they fall short. Learn to validate AI-generated code and recognize when human expertise is essential.
- Address Ethical Considerations: Explore good scientific practice when using AI in research. This includes transparency, reproducibility, data privacy, and broader concerns such as environmental impact and energy usage.
- BONUS: Integrate LLMs within Python Workflows: Learn to make API calls to large language models directly from Python, allowing advanced automation possibilities for your research.
- Bring Your Own Data: You can provide a dataset from your own research, and I will incorporate it as a working example throughout the workshop. This allows you to leave with analysis code directly applicable to your work.
Analyzing Qualitative Data with Generative AI Tools
Format: 1.5-2 days, online or on-site
This workshop introduces you to AI-powered approaches for qualitative data analysis—from transcription and coding to thematic analysis and automation. Whether you work with interviews, focus groups, open-ended survey responses, or other text-based data, you’ll learn how to leverage AI tools while maintaining methodological rigor.
What You Will Learn
- Semantic Analysis of Text-Based Data: Use AI tools to code and categorize qualitative data. Generate coding frameworks, identify themes, and systematically analyze patterns across interviews, focus groups, and open-ended responses.
- Automated Transcription Integrate transcription tools directly into Python workflows or use command-line solutions to efficiently convert audio to text, saving hours of manual work.
- Workflow Automation with LLM APIs: Learn to integrate large language models into Python scripts via API calls, enabling automated coding, batch analysis, and scalable qualitative research workflows.
- Navigate Limitations and Maintain Rigor: Understand the strengths and limitations of AI in interpreting context, nuance, and meaning. Develop strategies for ensuring analytical rigor, transparency, and trustworthiness in AI-assisted qualitative research.
- Choose Your Tools and Compare: Compare established qualitative analysis software (e.g., MAXQDA) with custom Python-based approaches. Understand when specialized software is beneficial and when flexible, code-based solutions offer advantages.
- Local and Institutional Models for Data Protection (BONUS): Explore privacy-preserving alternatives – running language models locally or using institutional deployments to analyze sensitive data without compromising confidentiality.
- Bring Your Own Data: You can provide qualitative data from your own research (interviews, transcripts, open-ended responses—appropriately anonymized), and I will incorporate it as a working example throughout the workshop. You’ll leave with analysis workflows directly applicable to your research.
Disclaimer: As you might expect, generative AI tools were used in parts of this workshop’s development, including brainstorming, text refinement, and image production. However, no full sentences or larger passages were generated without my (Alex Britz) direct input and oversight.