My research concerns developing androids (very human-looking robots) as a controlled experimental apparatus for testing hypotheses about human brains and behavior and as a testbed for theories about how neural or cognitive processes influence behavior. Results from my use of androids in experiments in human-android interaction have, for example, demonstrated cultural influences on how and why people break eye contact while thinking. My advances in developing more humanlike androids are being translated into a better understanding of the cognitive mechanisms underlying social interaction. This results in design principles for creating future generations of robots that are appealing and beneficial to humankind. My research is evidence-based, because the microdynamics of human-android interaction are statistically analyzed, as are the influences of androids on human attitudes, based on adequate samples sizes. It belongs to the field of android science, which is inherently interdisciplinary because androids are developed using technologies from artificial intelligence, pattern recognition, control systems, mechanical engineering, and material science, which are applied to deepening our understanding of social and cognitive psychology and cognitive neuroscience. The goals of android science are to better understand human beings, develop design principles for developing robots that support useful interaction with human beings, and developing robot applications that address the needs of our aging population.