The opportunities for research in the 1980s will be much affected by changes in society. In particular, by a clash between two powerful forces: the producer bureaucracy and the new individualism. It will be a clash of values and opinions and research will be needed to measure them. However, I believe there are dangers in the way that opinion research will be used, and especially in the way that it will be publicised. The protagonists in the arguments on many social issues are likely to make increasing use of published "research events" that is, over-simplified and superficial figures on public opinion, designed mainly to appeal to the media's need for simple and dramatic news, and thence to affect public policies. I think, in fact, that the evidence is very strong that there is a genuine difference between people's private opinions and their public opinions. But the requirements of media contests will tend to force research methods to concentrate heavily on public opinions. The danger is not so much that this would distort social policies (though it might) as that it could damage the reputation of market research in general, especially if, as is likely, the findings of one group's "research events appeared to be directly opposed to those of another's. One important challenge of the 1980s will be to identify these risks and do something to guard against them. As a start to the discussion, I have suggested five simple guidelines which I hope can contribute to a joint campaign to educate ourselves, our clients and the media in the proper use and presentation of opinion research.