Parametric Recovery Methods: A Comparative Experimental Study
We propose and implement an experimental methodology for comparing the predictive success of various methods for recovering individual preferences from choice data. We apply the proposed approach to a comparison of two parametric recovery methods: Non Linear Least Squares (NLLS) and the Money Metric Index (MMI). The former is based on minimizing the distance between observed and predicted choices while the latter is based on eliminating incompatibility between the ranking information encoded in choices and the ranking induced by the parametric specification. The experiment, in the context of choice under risk, involves a two-part design where choices made by subjects in the first part are used to construct their choice sets in the second part of the experiment in order to separate the predictions of the two recovery methods. We find that the Money Metric Index predicts better than NLLS in all cases and significantly better when the recovered parameters imply non-convex preferences.
|Date of creation:||09 Jan 2016|
|Date of revision:||03 Nov 2016|
|Contact details of provider:|| Web page: http://www.economics.ubc.ca/|
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
When requesting a correction, please mention this item's handle: RePEc:ubc:pmicro:yoram_halevy-2016-2. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Maureen Chin)
If references are entirely missing, you can add them using this form.