Faculty of Engineering and Natural Sciences
High-impact Research and Scientific Misconduct in Biomedical Sciences
Millions of active biomedical researchers are scattered around the globe, most focusing on a small fragment of the enormous biological universe. Consequently, knowledge acquisition by scientific communities is bound to have large-scale properties that are normally hidden from individual scientists. Using a huge data set which mirrors the knowledge growth in biomedical sciences, I will present several important features of this distributed thinking activity. These analyses suggest that high-impact research is characterized by popular molecular interactions coupled with novel interactions, and popular biological subjects coupled with unpopular chemicals and drugs. Second, I will describe a model of scientific misconduct, which estimates the magnitude of hidden false articles in the literature that should be retracted, and illuminates the relationship between visibility of research and self-correction of science. Finally, I will present SciTrends, a website for visualizing the evolution of biomedical fields over time.
July 31, 2007, 13:40, FENS G032