The purpose of this paper is to investigate the combined effects of different modalities of long-term knowledge accumulation and short-term knowledge searching on the generation of high-impact ideas. The authors aim at providing useful conclusions for academic scholars.
Two dimensions of the cumulative knowledge structures of researchers are measured: knowledge depth and knowledge breadth. The search strategies employed by researchers are classified as local search and distant search. The authors use researchers’ historical publications to measure cumulative knowledge structures. References contained in these publications serve as an indicator of knowledge searching behaviors and are used to measure search strategies. Highly cited papers with random-but-matched papers from the same journal published in the same year are adopted to test the hypotheses.
The knowledge depth of researchers positively predicts the generation of high-impact ideas. Knowledge breadth has a bell-shaped relationship to the generation of high-impact ideas. Two instances of “strategic fit” between the knowledge structures and search strategies of researchers are identified; namely, knowledge breadth combines most effectively with local search, and knowledge depth combines most effectively with distant search in generating high-impact ideas.
Using article references to measure authors’ knowledge search behaviors may lead to biases. Future research should perform a survey to obtain a comprehensive understanding of the knowledge search behaviors of scholars.
A “T-shaped” knowledge structure in the long run is recommended for maximal generation of high-impact ideas. Researchers who have not adopted this optimal knowledge structure can employ a matched search strategy to leverage their existing knowledge structures.
This paper is among the first to examine the interactive effects between the cumulative knowledge structures and short-term knowledge searching strategies of researchers. The authors have enriched the exploration and exploitation theory by adding the dimension of time into the analysis.
