This paper has been archived and removed from our list of current Technical Papers. The information it provides may be outdated or irrelevant based on our present understanding of the topic. However, we will continue to publish it here for historical purposes.
Version 8.1 of SSI Web includes an improved experimental designer for MaxDiff studies. This article describes how the new approach obtains superior results in terms of within-version item balance and quantifies the benefits if conducting individual-level analysis (HB). We assume the reader is already familiar with MaxDiff (Maximum Difference Scaling), also known as best-worst scaling, as well as HB estimation.