IF your idea of democracy is defending the right to complain, or to vote only when you sorta, maybe feel like it, and you never, never get involved in public issues because, c'mon, what good does it do, listen to Jim Fishkin. His vision of a different kind of citizen engagement will be seen and heard on the night of Jan. 18. Jim Lehrer will step in front of PBS television cameras to begin an extraordinary, three-part special on the National Issues Convention. There in the studio, Mr. Fishkin, a professor of government at the University of Texas, will have gathered some 600 United States citizens. They will be a carefully chosen representation of the American electorate. ''Even people who don't have phones,'' he says. If all goes well, most or all of the presidential candidates will be there, too. The objective is to deepen individual understanding of several key issues in the election year, to get away from the shallowness of ''poll-driven mass democracy, from sound bites and democracy by slogans,'' Fishkin says. He wants nothing less than to create ''an authentic public voice'' that knows the issues and can help restore genuine political dialogue. To achieve this, he will first have all 600 people respond at home to a National Opinion Research Center Poll about the issues. Then they all will be brought to Austin, Texas, to deliberate at length in small groups about the issues in the poll and to develop questions to ask the presidential candidates. Carefully balanced briefing materials will be prepared by the Kettering Foundation, an Ohio research organization with experience in organizing public forums. A bipartisan advisory committee will approve the materials. ''Much research has established that many of the opinions reported by polls are non-attitudes or pseudo-opinions,'' Fishkin says. ''These polls model what the public is thinking when it is not thinking, because respondents are often asked questions about which they have no knowledge or settled views.'' After separate televised meetings with Republican and Democratic candidates and more discussion, the citizens will be polled again to measure the changes and impact of the deliberative process on their thinking about the issues. A third PBS program will focus on the National Issues Convention itself to analyze the process for its wider implications. ''A deliberative poll models what the public would think if it had a better chance to think about issues,'' Fishkin says. ''In the groups, people learn from each other, from the dialogue.'' Fishkin's deliberative approach has been done successfully in Britain with 300 people, culminating in a nationally televised program on crime. The participants were polled. Then they deliberated in small groups and questioned law-enforcement officials, crime experts, and even criminals. Then they met with British political leaders on TV. The result of the process, according to Fishkin, was 300 people who, as a whole, remained tough on crime but grew to realize the limitations of prisons to stop crime and the complexity of the issue. They favored more rehabilitation efforts and wanted different treatment for first-time offenders. ''They became far more sophisticated consumers of the political parties' competing policies,'' Fishkin says. Fishkin proposed the idea of a deliberative poll in an Atlantic Monthly article in 1988. Since then he has garnered support slowly, raising a little more than $2.1 million with more commitments still pending, including a major airline to transport the participants. So far, major underwriters are the Federal Home Loan Mortgage Corporation, the city of Austin, and PBS. ''We're still in discussions with most of the candidates,'' says Fishkin, ''but we're hoping they can't resist the air time.'' PBS has committed to at least 11 hours of coverage. ''These are just ordinary citizens who will ask questions,'' Fishkin says, hoping the candidates will be up to the questions. Ervin Duggan, president of PBS, says the deliberative poll ''is the next generation of electronic town meetings and will add substance to election-year coverage.'' BUT will the intense discussion render the participants nonrepresentative of voters? ''They will have done something that we think most people haven't done,'' says Norman Bradburn, senior vice president of research for the National Opinion Research Center at the University of Chicago. ''But the conceptual purpose of this experiment is to determine if people will be more thoughtful because of the experience.'' After the deliberative poll in Britain, participants reported unique changes. ''One woman said that in 30 years of marriage, her husband had never read a newspaper,'' Fishkin says. ''But after being a part of the poll, he reads several newspapers a day.'' ''It's quite possible we underestimate the degree to which people already think about and discuss the issues,'' Mr. Bradburn says. ''The point here is not that the deliberative poll changes poll results, but does it change the behavior of people? Will they think about and discuss the issues more with their neighbors?''