Knowledge creation and transfer are conceptualized as part of a system of knowledge production that connects basic science with societal outcomes (such economic growth and national security). Hypotheses about knowledge creation are derived from the management of innovation and the organizational sociological literatures. Hypotheses about knowledge transfer, or more specifically gaps in transfer, are derived from the idea innovation network theory. Rather than viewing the knowledge production system as a single system, this study recognizes two primary sources of diversity: four different types of research projects defined by the strategic choices of the relative emphasis on normal science vs. high risk breakthroughs and small vs. large size projects; and the differences across scientific disciplines. Four critical methodological problems are addressed in this study. The first is to test for differences in the profiles of the four kinds of research projects as defined by a research environment survey that the researchers have developed and administered in several research organizations. The second is to attempt to measure innovativeness in real time with measures of technical progress and then correlate these with later measures of papers and patents. The third is to study the amount of technical exchange that is associated with various indicators (collaborations, joint papers) typically used in network analysis. The fourth is to measure the impact of cognitive distance on cross-functional teams and diverse functional collaborations. The potential impacts of this study are both intellectual and policy oriented. The findings would contribute to the management of innovation and organizational sociology literatures because concepts are defined in new ways and the research setting (science) is different. The most important intellectual contribution is the synthesis of these literatures with those on the idea innovation network and inter-organizational theory, which results in the recognition that the fundamental problem is the identification of gaps in the knowledge production system. The broader implication of this study rests not only on its contribution to a science of science and innovation policy platform but its contribution to a number of problems in the social sciences. The policy impacts are both immediate and long-term. The immediate ones are that descriptive reports are provided each participating department at selected national labs and center on 42 attributes of the research environment survey a few months after the data collection; a comparative report on the findings including differences by the type of the research project is provided at the end of the three year project. The long term impact is on the construction of econometric models of innovation in which the research organization is the micro unit. In particular, providing a way of measuring gaps in the knowledge production system is a critical policy issue as well as a new theoretical problem. Another contribution is the continued validation of the research environment survey for measuring innovativeness in different kinds of projects. A survey will be administered to all members of two departments and two centers at Brookhaven, Pacific Northwest, and Sandia National Laboratories, as well as one department and one center at Ames, the National Renewable Energy Laboratory, and the National Oceanographic and Atmospheric Agency. Further, a survey will be administered to 72 projects selected by middle managers at each of the laboratories. Projects will be selected that represent all four kinds of research projects defined by the survey, and project leaders will provide measures of technical progress to assess the surveys ability to assess innovation. Middle and top managers will report on mechanisms for encouraging integration across cognitive distances, strategies for the reduction in gaps in the idea innovation network and the characteristics of the five disciplines.
