In their paper they use a non-geometric design to generate the concepts. Non-geometric designs represents a class of orthogonal designs that when the assumption of effect sparsity is valid, i.e. that only a few of the attributes actually influence the respondents' preferences, provide an opportunity to analyze interactions between attributes as well as the attributes themselves. In this article the use of non-geometric Plackett-Burman designs for conjoint analysis is advocated. Also, a procedure based on restricted all subsets regression for taking advantage of the special characteristics of the non-geometric designs is proposed and demonstrated using data from a conjoint study performed on cellular phone antennas in Sweden. Blomkvist, Ekdahl and Gustafsson also conducted a Monte Carlo simulation to further illustrate the properties of the proposed procedure and the use of non-geometric Plackett-Burman designs for conjoint analysis