Article ID: | iaor20119373 |
Volume: | 8 |
Issue: | 3 |
Start Page Number: | 233 |
End Page Number: | 246 |
Publication Date: | Sep 2011 |
Journal: | Decision Analysis |
Authors: | Simon Jay, Guyse Jeffery L |
Keywords: | preference modelling, problem elicitation |
The equivalence of two elicitation methods (sequences and matching) has been assumed when empirically testing the traditional discounting model even though the respective literature has revealed results that are dependent on the procedure used. Three common anomalies revealed (gain/loss asymmetry, short/long asymmetry, and the absolute magnitude effect) are investigated using the two different methods in a within‐subjects experiment. In both procedures, it appears that the participants in this study evaluate monetary outcomes over time differently than the discounting model predicts. Patterns consistent with two of the anomalies (gain/loss and absolute magnitude effect) surface and interact in both elicitation techniques. Finally, a systematic inconsistency exists between the two methods. We observe significantly more consistency between the two elicitation techniques when the outcome is a gain in the relatively far future than when it is a future loss. This may be due to the participants' inability to display a preference for spreading losses, which they revealed in the sequences task, in the matching task.