Abstract
When analysing choice experiments respondents are assumed to attend all attributes and alternatives in the same way. However, because of the important role that the price of the alternatives can play in the decision-making process, the level of price of a specific alternative may have consequences on the level of consideration given to the remaining attributes of the alternative. In this article, we propose the use of a discrete mixtures logit approach to accommodate respondents ignoring alternatives in stated choice experiments. Our results indicate a higher propensity for respondents to ignore attributes when they face cheap rather than expensive alternatives. We also find that allowing for this leads to substantial improvements in the model performance.
Acknowledgements
The authors would like to acknowledge the comments of Fredrik Carlsson, Ric Scarpa and Peter Martinsson in earlier stages of the work described in this paper. Responsibility for any errors or omissions remain the authors'.
Notes
11 USD = 523 Chilean Pesos at the time of the survey. On average at the time the survey was fielded, the monthly electricity bill per household in Concepcion was approximately 12 000 Chilean Pesos.