Existential Analysis Codebooks for LLM-Assisted Qualitative Analysis
Two open-source codebooks supporting the WCET4 (2026) paper "Do You Understand?!" Best Practices using Artificial Intelligence in Research (and Life) by Graham Nelson-Zutter, MSc. Both are downloadable, citable, and licensed CC BY 4.0 for reuse in your own research.
A structured codebook for scoring participant testimony against Längle's Four Fundamental Motivations — the foundational framework. Built by multi-LLM cross-validation across Claude Sonnet 4.5, ChatGPT 5 Thinking, and Gemini 2.5 Turbo, using four primary Längle sources (1992, 2002, 2003, 2011). Every code traces to a verbatim Längle quote with citation.
A structured codebook for scoring participant testimony against Längle's Twelve Fundamental Existential Prerequisites — derived from the 4FM framework, decomposing each Fundamental Motivation into three Prerequisites for finer-grained scoring. Built by the same multi-LLM consensus methodology, against two primary Längle sources (2002, 2011b).
Each codebook is a JSON artifact that constrains an LLM to score participant testimony within Längle's framework, rather than letting the LLM draw on generic prior knowledge. They are not fine-tuned models and not prompts in themselves — they are structured references you provide to your LLM of choice alongside testimony.
instructions_for_llm / coding_guidelines block. The codebook requires the LLM to anchor every code in a verbatim Längle quote plus a verbatim testimony quote.The companion paper (below) walks through the three core techniques in detail — multi-LLM cross-validation, quote-anchored source verification, and iterative multi-pass with re-anchoring — that underlie both the construction and the application of these codebooks.
Nelson-Zutter, G. (2026, June). "Do You Understand?!" Best Practices using Artificial Intelligence in Research (and Life) [Conference presentation]. World Congress for Existential Therapy 4 (WCET4), Denver, CO, United States.
Methodology paper DOI: forthcoming (post-conference deposit).
Both codebooks were authored by Graham Nelson-Zutter, MSc student in Existential Analysis & Logotherapy at the University of Salzburg, in the course of his MSc thesis on LLM-assisted qualitative analysis of psychedelic-assisted therapy participant experiences.
The LLMs used in the methodology (Claude Sonnet 4.5, ChatGPT 5 Thinking, Gemini 2.5 Turbo, and others) are research instruments, not co-authors. See each codebook's README.md and docs/lineage.md for the full multi-LLM build provenance.