Supplementary data and generative workflow for "A Ranking-Based Human Validation Layer for Concept Implementation in AI-Generative Design"

This repository contains the supplementary materials associated with the paper "A Ranking-Based Human Validation Layer for Concept Implementation in AI-Generative Design," published in the Journal of Engineering Design (Taylor & Francis). The dataset includes: (1) the ComfyUI workflow used to implement the SDXL-based image-to-image generative pipeline; (2) the real product photographs used as base input images for each Product–Concept pair; (3) the prompt specifications defining the conceptual intensification across levels A–F; (4) the full set of AI-generated stimuli presented to participants; and (5) the anonymised ranking data and analysis scripts used to compute agreement metrics, Plackett–Luce aggregation, and NA-based operational range diagnostics. These materials are provided to support the transparency and replicability of the study. For further details on the experimental design and analytical procedures, please refer to the accompanying publication.