2004AgEcon Search (University of Minnesota, USA)Open access

USING THE RANDOM PARAMETERS LOGIT MODEL TO COMBINE REVEALED AND STATED PREFERENCE DATA

Brett R. Gelso, Jeffrey M. Peterson, Gelso, Brett R., Peterson, Jeffrey M.

Open full text 0 citations

Abstract

Recent literature has combined Revealed (RP) and Stated Preference (SP) data in the Multinomial Logit Model (MNL) to estimate the value of environmental goods. However, emerging research has identified that a limitation of the MNL is the assumption of Independently and Identically Distributed (IID) errors, resulting in inaccurate model predictions and inconsistent utility parameters. Our analysis applies an alternative method to combine RP and SP data that takes into account the heterogeneity in both the observable and unobservable components of utility. This allows us to test whether such heterogeneity has an important effect on predicting behavioral choices.

Open-access reader

About this research paper

What this paper is about

Recent literature has combined Revealed (RP) and Stated Preference (SP) data in the Multinomial Logit Model (MNL) to estimate the value of environmental goods. However, emerging research has identified that a limitation of the MNL is the assumption of Independently and Identically Distributed (IID) errors, resulting in inaccurate model predictions and inconsistent utility parameters. Our analysis applies an alternative method to combine RP and SP data that takes into account the heterogeneity in both the observable and unobservable components of utility. This allows us to test whether such heterogeneity has an important effect on predicting behavioral choices.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Recent literature has combined Revealed (RP) and Stated Preference (SP) data in the Multinomial Logit Model (MNL) to estimate the value of environmental goods. However, emerging research has identified that a limitation of the MNL is the assumption of Independently and Identically Distributed (IID) errors, resulting in inaccurate model predictions and inconsistent utility parameters. Our analysis applies an alternative method to combine RP and SP data that takes into account the heterogeneity in both the observable and unobservable components of utility. This allows us to test whether such heterogeneity has an important effect on predicting behavioral choices.

Key concepts: Unobservable, Multinomial logistic regression, Econometrics, Mixed logit, Revealed preference, Preference, Multinomial probit, Discrete choice

Related papers

Back to paper searchBrowse research topicsOriginal source
USING THE RANDOM PARAMETERS LOGIT MODEL TO COMBINE REVEALED AND STATED PREFERENCE DATA — Research Paper | ScholarLens