Human-centered Insight Discovery Method: Meta-Perspective Structure and Insight Discovery Map Show Promising Advantages Over Laddering in Expert Evaluation

Abstract

This study examines whether human-centered insight discovery tools can support deeper qualitative insights than conventional laddering interviews. We define insight for product and marketing decisions and propose two tools grounded in Personal Construct Theory: The Meta-Perspective Structure (MPS) for question design and the Insight Discovery Map (IDM) for analysis. Using minimal, length-matched method-only prompts, a large language model generated interview questions and analysis outputs for MPS/IDM and laddering across seven persona-topic scenarios. Seven experienced researchers conducted open-label, paired evaluations on Novelty, Usefulness, Depth, Clarity, and Persuasiveness. In this exploratory evaluation, results suggest that MPS/IDM tended to receive higher ratings on Novelty, Usefulness, and Depth, while Clarity was comparable across methods; Persuasiveness showed a favorable pattern for MPS. These findings suggest that person-first, overview-to-detail exploration and contradiction-explicit reasoning are promising for early-stage insight discovery. However, limitations include a small, non-blinded expert panel and reliance on LLM-generated outputs, warranting blinded human evaluation.