Optimal Two-Level Conjoint Designs for Large Numbers of Attributes

In this paper, we propose a simple strategy to construct D-, A-, G- and V-optimal two-level multi-attribute designs for rating-based conjoint studies. Our approach combines orthogonal designs and balanced or partially balanced incomplete block designs. In order not to overload respondents with complicated tasks, the designs hold one or more attributes at a constant level. The designs are variance-balanced meaning that they yield an equal amount of information on each of the part-worths. Some examples are provided to illustrate the method