Categories shape preferences: A model of taste heterogeneity arising from categorization of alternatives

We propose and test a choice model based on the notion that the category an alternative is perceived to fall into determines the attribute importance weights used to evaluate that alternative. For example, space is more important than fuel economy for an SUV, but the opposite is true for a commuter car. In our model, the weights associated with different categories are stable, but categorization decisions can be subject to significant context effects; that is to say, we suggest that context effects are due to task interpretation rather than to preference construction. We demonstrate that the model can correctly detect experimental manipulations of product categorizations, supporting its suitability as a tool to capture preference heterogeneity arising from categorizations outcomes. The model is then employed to analyze data from a discrete choice experiment and the results provide rich behavioral insights into how categorization influences choice processes.