Examining Interactions Between User Characteristics and Explanation Modalities on Inducing Complementarity

Achieving complementary performance in human–AI collaboration, where the combined efforts of humans and AI outperform either working alone, remains a significant challenge. Providing explanations for AI assistance is often considered a potential strategy to reduce human over-reliance on AI and enhance decision-making. However, empirical studies have shown mixed results regarding the impact of AI explanations on performance improvement. In this work, we extend this investigation by exploring an additional dimension: whether user characteristics influence the effectiveness of AI explanations in achieving complementarity. Using a geography-guessing task as the experimental setting, we find that user characteristics, such as openness and experience, interact with explanation modality in inducing complementarity. Our results suggest that tailoring explanations based on user characteristics could enhance complementarity and provide insights into how personalized AI explanations can improve human–AI team performance.