Overview

Recent AI developments have opened new perspectives for qualitative research. While corpus-based discourse or content analysis traditionally treated texts as a "bag of words" and identified linguistic patterns based on frequency distributions, Large Language Models (LLMs) offer a significantly more profound ability to process language, semantics, and context.

This workshop will introduce how LLMs can be integrated into qualitative research by providing insights into their capabilities and limitations, alongside hands-on training in prompt engineering, model implementation, and critical bias evaluationโ€”all demonstrated through practical analysis of earthquake newspaper reporting and their argumentative units.

September 10, 2025

Session 1

Using LLMs for Qualitative Data Research

09:00-10:30

Presentation: Potentials and limitations of LLMs for qualitative research; Responsible Usage of LLMs.

Session 2

Model Implementation - Using Open Source/Weight Models

11:00-12:30

Hands-On: How to use open weight or open source models via JupyerHPC, Huggingface Jobs and API for local models.

Session 3

Qualitative Data Research - From Unstructured to Structured Data

14:00-15:30

Learn how to apply qualitative research methods via LLMs and transform unstructured data into structured data. Learn how to prompt engineer for qualitative research methods.

Session 6

Free Practice Session: Create Pipeline for Your Own Research

16:00-17:30

Free Practice Session: Create your own project pipline and try out models and computing environments with your own data.

September 11, 2025

Session 5

Quality and Bias Evaluation

09:00-10:30

Critical assessment of LLM outputs, understanding biases, and implementing quality control measures in research workflows

Session 4

Using Small Models for Qualitative Research - Fine-Tuning Models

14:00-15:30

Let's experiment: can we improve results with small models through fine-tuning?