The Homerun Hypothesis : Influencing the Boundaries of Knowledge

We argue that the creation of new knowledge is both difficult and rare. More specifically, we posit that the creation of new knowledge is dominated by a few key insights that challenge the way people think about an idea; generating high interest and use. We label this the homerun hypothesis. Using a sample of 2,012 published management studies we find support for the homerun hypothesis. Our evidence suggests that new knowledge creation is rare among the leading journals in management. We also find that numerous studies in the leading management journals are virtual strikeouts, making no contribution to knowledge. Additional tests indicate that journal "quality" is related to the ratio of homeruns to strikeouts a journal publishes and that journal rankings are a poor proxy for study influence. Consistent with the notion that editorial boards are able to identify new knowledge, we find that research notes significantly under-perform articles in both the same journal and in lower ranked journals. Taken together, the results imply that only a few scientific studies, out of the thousands published in a given area, change or influence the boundaries of knowledge, with many of the others contributing nothing. Overall, this analysis indicates that the development of new knowledge is rare yet appears to be recognizable. A revised version of this paper, titled "The Blockbuster Hypothesis" is forthcoming in the journal Scientometrics