There is no recipe for involving users but six suggestions can be offered. Individuals would have to adopt the attitude these suggestions imply, but full implementation would involve institutions.
First, expect resistance and difficulty. By definition users are not primarily interested in science---they have other concerns, and may see science as a distraction, a hoop to be negotiated, or just another political tool to be manipulated [Collingridge and Reeve, 1986; Fumento, 1994]. Also some users literally will not have been born yet. Thus it will not be easy to find and work with an adequate set of users, and reasonable proxies must be sought. Congress may serve as one such proxy. The modern question, ``who are your customers?'' is non-trivial.
Second, the idea that users are only a nuisance is counterproductive. This attitude ruled out early user involvement in NASA's EOS program. Eliminating it would be a culture change, effectively giving increased value to applied research. The attitude may be based in part on fear of letting users, especially politicians, manage science, but this fear is overblown because science per se does not interest politicians. They want only to lay out areas of need, described and defined in terms of social parameters, leaving a great deal of latitude for scientific choice.
Third, plan and carry out a variety of small-scale efforts to involve users at all levels, in all social and economic sectors and geographical regions. A variety of approaches is necessary, as no single approach will work for all situations. It would be important to evaluate all efforts, report what works, and propagate those [see e.g. Brunner, 1993]. Reinstituting congressional fora to bring together scientists and policy makers (such as those formerly organized by the University Corporation for Atmospheric Research) would be a start. Local efforts would be important also.
Fourth, sincerely attempt to incorporate user needs into research plans. This is another cultural change, effectively adding new external criteria to the direction of science and also abandoning the Bush model of automatic delivery of benefits. It will also need institutional change, such as setting up mechanisms for meaningful communication between scientists and users. Thoughtful interpretation of the user input will be needed because it will not be in the scientific idiom. For example, it will require thinking through connections between science and social problems, especially in areas characterized by ignorance where fundamental research may be the best approach. This also will require scientists to evaluate the relevance of their work and, more difficult, to police the budget justifications of funding agencies where some egregious exaggerations of usability occur. This policing might best be done through independent agency and National Academy of Sciences advisory committees.
Fifth, keep expectations modest. Democratic decision processes, especially the legislative process, are fitful, chaotic, incoherent, raw, and patched---in a word untidy---partly because they resolve differences of values and interests. We keep these processes because we consider them and their results valid. Involving users will be a slow, untidy process but it will increase the untidiness of research only marginally because frontier research is rarely cut-and-dried. Scientists and users, through mutual involvement, may even come to understand each other's untidiness. For Congress to gain a thorough understanding of science would be valuable to both institutions.
Finally, consider the whole citizenry as occasional users of science who could become more involved in its support. There is broad public interest in and support for science, but it tends to be passive and woefully uninformed. To change this will require more science learning, not just teaching. (note: If this distinction is not clear, recall the story of the man who was teaching his dog to talk. When a friend asked to hear the dog speak the man said ``Oh, he can't talk, I'm just teaching him.'' As scientists deal with naive users of science they might keep one thing in mind: Many users they deal with, e.g., government officials, are college graduates who rejected the science teaching offered them.) How to produce scientifically literate citizens is not clear, but surely it is difficult, as many sincere, well-funded, but only moderately successful efforts testify. A first step in improving undergraduate science learning might be to grade and reward university science departments based on a test of critical thinking given all graduating seniors, not just the science majors. After all, each is a citizen who will become an alumnus or alumna, taxpayer, voter, official, maybe even a Congressional appropriations subcommittee chair.