Back to news
Ethics Society

A new tool makes the joint interpretation of humans and ChatGPT visible in qualitative analysis

The research introduces a new methodological tool that allows for step-by-step tracking of how meanings are constructed when an analyst interprets data in dialogue with generative AI.

In an article published in the AI & SOCIETY journal, María Paz Sandín Esteban describes an instrument called A Duo with ChatGPT. Its goal is not to assess how 'well' ChatGPT performs, but to make the so-called hybrid interpretation empirically traceable: when the interpretation is done by a human, when by the system, and how these phases intertwine in a qualitative analysis task.

The tool is based on a functional definition of hybrid interpretation and structures the phenomenon through four cognitive dimensions. The first is shared agency: who guides the interpretation and decisions at different points. The second is interpretive abduction, reasoning where the best explanation and hypotheses are formed based on observations. The third is cognitive control: how the analyst maintains a grip on which claims and conclusions are accepted. The fourth is analytical metacognition, practically the examination and evaluation of one's own thinking during the analysis.

The instrument organizes the analysis task into sequential phases, alternating between independent human interpretation and guided interaction with ChatGPT. This creates a structured 'trace' of how the interpretation progresses and what roles the human and generative system assume at different moments.

Source: Hybrid interpretation with generative AI: a pilot study using the A Duo with ChatGPT instrument, AI & SOCIETY.

This text was generated with AI assistance and may contain errors. Please verify details from the original source.

Original research: Hybrid interpretation with generative AI: a pilot study using the A Duo with ChatGPT instrument
Publisher: AI & SOCIETY
Authors: María Paz Sandín Esteban
January 13, 2026
Read original →