Bridging the Dual Nature: How Integrated Explanations Enhance Understanding of Technical Artifacts

Integrated explanations improve practical ability more than focused ones: When explanations combined both how something works internally (Architecture) and why it matters (Relevance), participants were significantly better at actually using their knowledge, even though single-focused explanations conveyed equivalent factual understanding. This suggests that for responsible AI, users need both technical details and context about purpose to truly engage with systems.

Understanding "how" requires bridging two perspectives: The study found that comprehension alone—knowing that something works a certain way—can be achieved through either architectural or relevance-focused explanations. However, the ability to do something with that knowledge (enabledness) specifically required integrated explanations, indicating that responsible AI design must deliberately connect technical and contextual information rather than treating them separately.

Current AI explanations may leave users unprepared for real-world use: The significant difference in "enabledness" between integrated explanations and focused ones (F(1,102)=4.83, p=.030) suggests that many existing AI explanations—which often emphasize either technical mechanics or business purpose in isolation—may leave users unable to effectively apply their understanding in practice, potentially undermining informed decision-making about AI systems.

Technology education and AI transparency need redesign: The findings indicate that explainable AI (XAI) systems and educational materials should be deliberately structured to connect internal workings with real-world relevance, rather than presenting these as separate topics. This has direct implications for how organizations should communicate about AI systems to stakeholders and the public.