Removing response-style effects in attribute-determinance ratings to identify market segments

Previous research has indicated that row centering or standardization will remove response-style effects in attribute-determinance ratings. Cluster analysis solutions based on raw and individually standardized attribute-determinance ratings are generated and compared for a major energy-consuming household durable. The standardized data solution produces clearer and more meaningful market segments and is more strongly and significantly related to a number of purchase-related variables, as well as anticipated purchase intent to hypothetical conjoint-analysis alternatives. The raw data solution suggests that response-style effects are strongly present and carry over to other types of rating scales. It is recommended that researchers remove response-style effects and attempt to capture product-class importance when using attribute-determinance ratings as inputs to cluster analysis to identify market segments.