Lecture notes “AI in qualitative research: Between acceleration, erosion and methodological innovation”
Updated script of the lecture by Dr. Thorsten Dresing from 05.02.2026 (PH Tirol)
You can download the script of my presentation here. On around 20 pages, it documents my personal situation in February 2026 and shows The changes brought about by AI are no longer hypothetical. Tools such as Claude Co Work enable non-programmers to develop customized research software in hours. A systematic re-analysis experiment proves that an automated qualitative evaluation, which originally required 100 working hours, can probably be reduced to a fraction of the time with iterative LLM procedures – with correct citation and content comparability with the manual analysis. At the same time, the problem is getting worse: six recent studies paint a worrying picture of cognitive consequences. Over 90% of students use AI tools, neuroscientific findings show minimal brain activity during AI-supported work, and even highly specialized experts lose over 20% of their interpretation skills after just three months of AI assistance. However, structured, reflective use with didactic support proves to be an effective countermeasure.
The script focuses on six methodological approaches that have been published or are in preparation – from the documentary method with AI (Schäffer/Lieder) to extended qualitative content analysis (Kuckartz/Rädiker), query-based analysis (Morgan), conversational analysis with AI (Friese) and hybrid interpretation and serendipity prompting (Krähnke/Pehl/Dresing). These approaches are united by the aspiration to use AI as a multi-perspective sparring partner rather than an automated substitute – and they show that methodological innovation is possible without compromising scientific standards.
The script presents GDPR-compliant implementation paths: from the GWDG Academic Cloud to local processing with open-weight models and specific hardware recommendations.
The transformation is taking place. Whether it leads to the erosion or enrichment of qualitative research competence depends on whether we actively help to shape it.
Current script (February 2026)
PDF version of the script, based on the automatically transcribed and AI-edited lecture from 05.02.2026 at the PH Tirol.
Earlier script (October 2025)
The first version of this script – in parts already outdated, but worth reading as a document of the rapid development. A comparison of the two versions makes it clear what new findings and changed assessments have emerged in just a few months. PDF version of the script based on a transcript of the lecture from 17.10.2025.
