Generation and Impact of Novel Articles in Physics

The aim of this paper is twofold. First, it investigates how the characteristics of a research team relate to the probability of generating a novel article. Second, once a novel article has been generated, it investigates the moderating effects of the team's characteristics on the impact of the article. We consider 42,493 teams of researchers publishing a corresponding number of articles in 273 reputed journals in physics from 2005 to 2009. We find that team experience and team specialization are negatively associated with the probability of generating a novel article. On the contrary, having already written novel articles in the past is positively associated with the probability of generating a novel article. When analyzing the impact of novel articles, we find that a novel article published by an experienced team receives fewer citations than a novel article published by a team with less experience and is published in a lower impact factor journal. We also find that a novel article published by a large established team of researchers receives more citations than a novel article published by a small newly formed team. The team size is also positively related to the impact factor of the journal in which the novel article is published.