Discussion on the Invited and Contributed Papers

The paper of Drs Horvitz, Greenberg, and Abernathy, and the other papers which Dr Greenberg has organized for this session will surely make the record of this session a key reference for all future developments of the theory and practice of randomized response. We must be most grateful to them for their most clear, concise and thorough summary of developments to date. We are all very glad to see the originator of randomized response, Dr Warner, here today, and one must agree with his plea for a generalized approach to the issues. In that connection I have two comments to make. First, Dr Warner's neighbour in Toronto, D. A. S. Fraser, has developed a theory of statistical inference in which one has the formal representation