It might be expceted that a more active conversation betwecil cnumerator and respondent would provide better data in an interview. Such an effect had been report!:d for health interviewing. However, it was not discovered in farm intervicwing in the survey analyzed in this paper. There is probably an effcct on the quality of information from an active conversation interaction but it is likcly to be a complex one that requires in-depth study to describe.