Course on Structural Modeling

Dissertation of the year

Research Organizations

NOBEM Organization

PREBEM

Links

Jobs

Course on Structural Modeling

Michel Wedel

Groningen

January 8^th and 9^th 2007

The objective of this course is to provide an overview of research in structural models in marketing, particularly as it regards the New Industrial Organization paradigm (NEIO). The goal of this course is not only to introduce a body of literature on a rapidly emerging field but also to give you a strong foundation in critical thinking in the modeling area and to help you develop your own research interests. The fact that most of the papers discussed in the course are very recent testifies to the speed with which this area develops.

You are expected to read all the papers (10 papers) and evaluate them critically. For every session, read the papers in the order that they are listed in the syllabus. More specifically, when reading the papers think about the following issues for each paper.

1.	Why do you think this paper was or should be published? What does it contribute to our knowledge?

2.	What would you have done differently in writing the paper or doing the research?

3.	Are there any weaknesses in the paper? What are critical assumptions on which it is based? Are these defensible? Are there alternative approaches? Does it contribute to what we knew already?

4.	What may be future research directions? Can the theoretical framework be used to understand problems in consumer behavior and marketing strategy?

For every session, read the papers in the order that they are listed in the syllabus. All papers are available in pdf format. There is also a list of recommended readings. We will not extensively discuss these papers in our meetings but they are included them in the syllabus because they are important papers in the areas that we cover.

Before the sessions, you have to hand in assignments A to H. After the entire workshop, you need to write a short proposal about your own research as a result of this course (1-5 pages).

Day 1, January 8, 2007, Part 1, 9-12:

Review of Structural Modeling

In the first meeting, we will discuss issues of the review process. Please read the following articles before our first meeting. We will talk about their content and about writing/publishing and reviewing papers.

Chintagunta, P. K., V. Kadyali, N. Vilcassim. 2004. Structural models of competition: A marketing strategy perspective. C. Moorman, D. Lehmann, eds. Assessing Marketing Strategy Performance. The Marketing Science Institute, Cambridge, MA.

Chintagunta, P.K., T. Erdem, P. Rossi, M. Wedel, 2006. Structural Models in Marketing: review and assessment. Marketing Science, Anniversary Issue, forthcoming.

Recommended:

Reiss, P. C., F. A. Wolak. 2002. Structural econometric modeling: Rationales and examples from industrial organization. Handbook of Econometrics forthcoming.

Assignment

A.	Read the two papers in the order presented. Explain whether you think this is a good paper, and why. Are there any flaws in the paper? Does it explain the material well? Is it well written? Does it provide (new and relevant) insights? What are the main insights?

Day 1 January 8, 2007, Part 2, 1-3 pm:

Models of Consumer Choice

Guadagni, P. M. and J. D.C. Little 1983. A Logit Model of Brand Choice. Marketing Science, 2 (Summer), 203-238.

Train, Kenneth. 2003. Mixed Logit. Chapter 6, Discrete Choice Models with Simulation. Cambridge University Press.

Recommended:

Kamakura, W.A. and G.J. Russell 1989. A Probabilistic Choice Model for Market Segmentation and Elasticity Structure. Journal of Marketing Research, 26, 339-390.

Allenby, G.M. and P.J. Lenk 1994. Modeling Household Purchase Behavior with Logistic Normal Regression. Journal of the American Statistical Association, December, 89, No. 428, 1218-1231

Allenby, G.M., N. Arora and J.L. Ginter 1998. On the Heterogeneity of Demand. Journal of Marketing Research, August 35, 384-389.

Assignments

B.	Explain how the Train (2003) Chapter extends the original Guadagni & Little (1983) paper. Can you show, based on the likelihoods of the models in question, if and how the latter two give rise to G&L as a special case? What is the contribution? What are limitations?

C.	What is your view on the appropriateness of the form of heterogeneity proposed by Train?.

Day 1 January 8, 2007, Part 3, 3-5pm:

Types of Individual Differences

Wedel, M. W.A. Kamakura , N. Arora, A. Bemmaor, J. Chiang, T. Elrod, R. Johnson, P. Lenk, S. Neslin, C.S. Poulsen 1999. Discrete and Continuous Representations of Unobserved Heterogeneity in Choice Modeling, Marketing Letters, 10 (3), 219-232,

Andrews, R. Ainslie, A., Currim, I.S. 2002. An Empirical Comparison of Logit Choice Models with Discrete Versus Continuous Representations of Heterogeneity. Journal of Marketing Research, 39(November) 479-487.

Recommended:

Allenby, G., P. Rossi. 1999. Marketing Models of Consumer Heterogeneity. Journal of Econometrics 89 57-78.

Hutchinson, J.W.., W.A. Kamakura, and J.G. Lynch, Jr. 2000. Unobserved Heterogeneity as an Alternative Explanation for “Reversal” Effects in Behavioral Research. Journal of Consumer Research, 27 (December), 324-344.

Assignment

D.	Read Wedel ea (1999) first. Do not read the other papers for now. Think about your views on the opinions voiced in this paper. Design a study that would provide an answer to these questions, assuming you would be able to estimate the models in question.

E.	Now read the Andrews ea (2002) paper and compare with your answer in part D. What would you have done differently to test the differences between these methods?

Day 2, January 9, Part 1 9-12am:

Estimation of Individual Choice Models from Aggregate Data

Nevo, A. 2000. A Practitioners Guide to Estimation of Random-Coefficients Logit Models of Demand. Journal of Economics and Management Strategy, 513-548.

Bodapati, A. and S. Gupta 2004. The Recoverability of Segment Structure from Store-Level Aggregate Data. Journal of Marketing Research, 41, August, 351-364.

Recommended:

Draganska, M., D. Jain. 2004. A Likelihood Approach to Estimating Market Equilibrium Models. Management Science 50(5) 605-616.

Assignment

F.	Discuss the magic of estimating disaggregate models with individual differences from aggregate data. Do you really believe this can be done? Try and come up with at least two points of critique to the method described by Nevo (hints: see the Chintagunta Review in Session 1).

G.	Compare frameworks for heterogeneity used in each paper as well as the estimation method and provide your comments.

Day 2, January 9, Part 2, 1-5pm:

Bringing In the Supply Side

Kadiyali, V., K. Sudhir, V. R. Rao. 2001. Structural Analysis of Competitive Behavior: New Emprical Industrial Organization Methods in Marketing. International Journal of Research in Marketing 18 (1-2) 161-186.

Berry, S.T., 1994. Estimating Discrete-Choice Models of Product Differentiation. RAND Journal of Economics, 25, 242-262.

Recommended:

Yang, S., Y. Chen, G. Allenby. 2003. Bayesian Analysis of Simultaneous Demand and Supply. Quantitative Marketing and Economics 1 251-275.

Berry, S., J. Levinsohn, A. Pakes. 1995. Automobile Prices in Market Equilibrium. Econometrica 63 841-890.

Assignment

Delineate the assumptions that are made on the supply side in Kadiyali (2000). Describe how Berry (1994) generalizes Nevo (2000) in session 4. What is the main insight in Yang ea (2004)? Provide one way in which it is a step forward from Berry, and one way in which it is a step back. Provide a number of issues that need to be addressed in future research. Do you have ideas for future research?

Supplementary: Try and read “BLP”, it’s the paper that spawned much of the interest in the NEIO field in marketing.