Impact of image-generation tool on designers’ cognitive styles and design performance during product conceptual design

Image-generation tools have been widely adopted in product conceptual design, showing promise in supporting cognitive activities and fostering creativity. However, existing studies have primarily focused on the aggregate-level impact, overlooking individual differences in designers’ cognitive styles, making it difficult for AI to understand and adapt to the needs of different groups. To fill this gap, we aim to explore whether and how image-generation tools affect designers’ cognitive styles and to analyze the relationship between cognitive style and design performance. Through a within-subject experiment, we invited 28 designers to complete conceptual design tasks under SK-condition (traditional sketching) and AI-condition (image-generation tools). We found that AI-condition significantly influenced designers’ cognitive styles, with approximately 50% of designers exhibiting cognitive style shifts. Although changes in cognitive style did not have a significant impact on design performance, there was a significant association between cognitive style and design performance under AI-condition. Designers identified as Treasure Hunters performed best in novelty, variety, and fluency, while those exhibiting Focused Prober style, traditionally associated with experts, performed worse under AI-condition. These empirical findings provide valuable insights for customising tools and educational approaches that align with the needs of different cognitive styles.