The heterogeneous best-worst choice method in market research

Susana Tavares

Caixa Econmica, Montepio Geral

Margarida Cardoso and José G. Dias

ISCTE - Lisbon University Institute


The level of importance that consumers attach to the attributes of a product/service has been pivotal in understanding the purchasing decision process (Louviere et al. 2001). In order to measure the relevance of those attributes, there are two main approaches: the direct and the indirect one. In the direct approach, respondents are asked to state their valuation of a given attribute, usually by grading the attributes on a metric scale (rating scale) or by allocating points to the attributes (e.g. 100 points) (constant sum scale). The direct method (DM), based on rating scales, presents well-known disadvantages (Cohen & Markowitz 2002), namely:

  • rating scales do not provide enough diversity in the patterns of responding, in particular with short-amplitude scales (e.g. a 5-point scale)
  • different respondents tend to prefer different parts of the scale, either providing very homogeneous ratings or, on the contrary, the extreme points to indicate a higher or a lower relevance
  • in multi-country studies, cultural differences are generally observed in scale usage by respondents (Baumgartner & Steenkamp 2001).