The University of Leeds and Procter & Gamble Technical Centres Limited KTP 23_24 R6

A new statistical method borrowed from educational testing is being used to predict how people *feel* about a product—like a shampoo or detergent—based on how well it actually cleans. This matters because companies currently rely on large-scale consumer surveys and focus groups to link product performance to emotional response. These methods are slow, expensive, and often imprecise. The gap is a reliable, efficient way to measure the relationship between objective technical metrics—such as how much grease a soap removes—and subjective feelings like “this makes my hair feel soft.” The researchers are applying Rasch methodology, a technique originally developed to measure abilities in education (e.g., test difficulty vs. student skill), to model consumer attitudes with the same mathematical rigour. If this succeeds, manufacturers could replace sprawling consumer panels with smaller, faster tests that still yield statistically robust predictions. The immediate impact is on product development cycles: a company could tweak a formula and know, within days rather than months, whether consumers will perceive it as an improvement. This is applied research with a direct commercial endpoint—it does not aim to uncover fundamental principles of human perception, but to make an existing industrial process more efficient.