Introducing the fusion innovation test as a new paradigm for studying realworld creative problem solving.

Studying creative problem-solving requires tasks with real-world relevance and minimal domain-specific knowledge. The widely used alternative uses test (AUT) meets these criteria but primarily measures divergent thinking-generating many ideas without immediate judgment. However, creative problem-solving also requires convergent thinking to narrow down options and select the most effective solution. We introduce the fusion innovation test (FIT), a task that integrates both divergent and convergent thinking. In the FIT, participants combine two randomly assigned elements (e.g., objects or technologies) to achieve either a self-improvement goal or a sustainable development goal. Each FIT solution is evaluated based on the novelty and feasibility of combining the elements, and the effectiveness in achieving the intended goal. We collected behavioral data from 144 participants and used ratings from 122 human evaluators to train an automated scoring system based on a large language model. Participants also completed the AUT and two real-life creativity questionnaires. Results show that the FIT captures a unique aspect of creativity-goal-directed innovation-not assessed by the AUT. Moreover, FIT performance correlates with scientific creativity, whereas AUT performance aligns more with artistic creativity. These findings suggest that the FIT is an initially promising and complementary tool for studying real-world creative problem-solving.