nological dimensions. The social factors include such considerations as commitment, collaboration, communication, the treatment of multiple viewpoints, workplace diversity, and the willingness to take risks. Administrative factors include salaries, benefits, workplace conditions, the availability of sabbaticals, and travel funding. Technological factors include technical support, training, access to high-speed computing and communications, information services, and so on.

Each of these environmental and cultural dimensions can itself be the subject of innovation. This is most obvious with regard to information technology. To take just one example, a Web site called InnoCentive (www.innocentive.com) now allows companies to post R&D problems online and offer scientists financial rewards for solutions.

The consequences of innovation extend into the social and administrative spheres. For example, increasing the number of women in the biomedical sciences helped focus attention on women’s health issues, with corresponding increases in research in these areas. Similarly, funding researchers at different stages in their careers and at different types of institutions can expand the range of viewpoints brought to bear on a problem.

The federal initiatives that could improve the research environment and culture are unlimited. Among those suggested are the following:¹

• Increase the size and duration of project awards so that researchers spend more time doing research and less time ensuring that their research is supported (see Figures RP-1 and RP-2).

• Increase the diversity of the individuals and organizations doing research.

• Fund risky projects that could dramatically advance an area of research or open new research frontiers.

• Develop a new digital cyberinfrastructure to make the best use of rapidly expanding databases and multidisciplinary collaborations.

• Expand funding for merit-reviewed, cross-disciplinary, collaborative research centers.

• Collect the best practices and attributes of federal agencies and research performers and disseminate this information widely.

• Develop a common electronic grant-application system that combines the best features of current systems and can be used by all researchers and all federal agencies.