To collect valid and actionable information for positioning and product development of neuroleptics in case of schizophrenia, 420 psychiatrists and 60 schizophrenic patients were interviewed using a combination of multi attribute models (conjoint analysis and perceptual mapping). Because the trade-offs were based on non-discrete attributes, the market simulation had to be based upon the perception of the neuroleptics instead of the a priori definition of the neuroleptics. The results of both surveys indicated the importance of the improvement of the quality of life. At the same time, the results indicated the feeble performance of classic neuroleptics in this respect. The experiment of interviewing schizophrenic patients and letting them administer a series of trade-offs generated very useful results and showed that a common frame of reference and additional information to interpret the results of each individual is a prerequisite. This conclusion does not only hold for the interviewing of schizophrenic patients, but seems valid for a wide variety of conjoint studies. To interpret conjoint results and the results of multi- attribute models, one has to understand the frame of reference behind the answers. It is only with this additional information that we can determine the attributes that trigger the response.