Data and code for "The illusion of objectivity: affective prompt framing and apparently neutral AI advice in human–AI judgment"

This repository contains the data and code supporting the study “The illusion of objectivity: affective prompt framing and apparently neutral AI advice in human–AI judgment.” It includes the original Study 1 YAML model outputs, cleaned computational and de-identified behavioural datasets, prompt templates, corpus traceability and integrity records, statistical analysis scripts, figure-generation code, generated analytical outputs, and detailed reproduction instructions. The deposited original model outputs permit reproduction of the reported analyses without renewed access to commercial APIs. Optional regeneration scripts require separate provider credentials and may produce different responses because hosted models and APIs change over time. No directly identifying participant information is included.