Integrated mixed logit and latent variable models
AbstractA traditional discrete choice model assumes that an individual's decision-making process is based on utility maximization and that the systematic part of the utility function depends on some observable attributes and covariates. These attributes and covariates however can only explain part of the utility and a large part remains unexplained. In recent years, researchers have recognized that psychological factors such as attitudes, lifestyle and values can affect an alternative's utility and hence the individual's choices. Therefore, extending the traditional discrete choice model by incorporating those latent or unobservable factors, can help to better understand the individual's decision making process and therefore to yield more reliable part-worth estimates and market share predictions. Several integrated choice and latent variable (ICLV) models which merge the conditional logit model with a structural equation model exist in the literature. They assume homogeneity in the part-worths and use latent variables to model the heterogeneity among the respondents. The current research uses the mixed logit model that describes the heterogeneity in the part-worths and uses the latent variables to decrease the unexplained part of the heterogeneity. The empirical study that we conducted to assess student preferences on mobile phone features shows these ICLV models perform very well with respect to model fit and prediction. Furthermore, we compare the different estimation procedures that exist in the literature. Results show that, as expected, the simultaneous procedure where we estimate the structural equation model (SEM) and the choice model simultaneously, gives better model fit and more accurate predictions compared to the sequential procedure that estimates the SEM first and then the choice model taking the estimated factor scores into account. But the gain of the simultaneous procedure is relatively small. Therefore one can use the easier sequential procedure without losing much efficiency if needed.
Download InfoIf you experience problems downloading a file, check if you have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.
Bibliographic InfoPaper provided by Katholieke Universiteit Leuven in its series Open Access publications from Katholieke Universiteit Leuven with number urn:hdl:123456789/347872.
Date of creation: May 2012
Date of revision:
Contact details of provider:
Web page: http://www.kuleuven.be
Discrete choice models; Structural equation models; Hierarchical Bayesian estimation; Panel mixed logit model; Heterogeneity distribution;
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.:
- Asim Ansari & Kamel Jedidi, 2000. "Bayesian factor analysis for multilevel binary observations," Psychometrika, Springer, vol. 65(4), pages 475-496, December.
- Bliemer, Michiel C.J. & Rose, John M., 2010. "Construction of experimental designs for mixed logit models allowing for correlation across choice observations," Transportation Research Part B: Methodological, Elsevier, vol. 44(6), pages 720-734, July.
- Arora, Neeraj & Huber, Joel, 2001. " Improving Parameter Estimates and Model Prediction by Aggregate Customization in Choice Experiments," Journal of Consumer Research, University of Chicago Press, vol. 28(2), pages 273-83, September.
- Toubia, Olivier & Hauser, John & Simester, Duncan, 2003. "Polyhedral Methods for Adaptive Choice-Based Conjoint Analysis," Working papers 4285-03, Massachusetts Institute of Technology (MIT), Sloan School of Management.
- David J. Spiegelhalter & Nicola G. Best & Bradley P. Carlin & Angelika van der Linde, 2002. "Bayesian measures of model complexity and fit," Journal of the Royal Statistical Society Series B, Royal Statistical Society, vol. 64(4), pages 583-639.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Carl Demeyere).
If references are entirely missing, you can add them using this form.