Where do breakthrough ideas come from? Characteristics of scientists´ research behaviors

Small fraction of scientists garnering the majority of citations. In this paper, we compared “highly-cited scientists” from four disciplines with random-but-matched scientists from the same institution, same domain and with the same title. We find that explorative research is curvilinearly and exploitive research positively related to the occurrence of breakthrough ideas. Scientists´ heterogeneity of prior knowledge strengthens the curvilinear relations.