26:630:675 Marketing Models
Rutgers Business School – Newark and New Brunswick
Spring 2015
Instructor: Xiaowei Xu
Email: xiaoweix@andromeda.rutgers.edu
Lecture: Tuesdays, 9:00-11:50AM
Location: WP402
Office: 100 RR, Rm 3141
Tel: 848-445-9224
Office Hours: by appointment
Goal: This course provides broad statistical knowledge and basic programming skills for
conducting quantitative marketing research projects. Causality between sales/market
shares/customer purchases and marketing mix (e.g., price, advertising, and promotion
activities) is an essential logic link in many marketing models. Statistical causality is
often established with generalized linear models (e.g., linear/nonlinear, logit, multinomial
and Poisson regressions). This course covers the basic theory of GLMs and empirical
applications in market response models. The other part of this course includes: Bass
diffusion models, empirical research on pass-through rates, and hazard rate models (if
time permits). All data analysis is implemented with R.
Prerequisite: Elementary statistics and computer skills; Intermediate microeconomics.
Statistical packages: R (free download and tutorial at http://www.r-project.org/)
Related Courses: 26:799:685 Supply Chain-Marketing Interface
Class Participation: Class participation is mandatory. Students are required to discuss
class materials in class.
Homework: Students are required to analyze the in-class datasets in depth and
demonstrate their results in class.
Presentation: Two students form one presentation team. Each team is required to give
two presentations based on journal papers (see details on Page 4).
Project: Two students form one project team. Each team is required to write a mini
marketing research project and present the project results in class (see details on Pages 23).
Exam: Students are required to analyze datasets and build market response models using
R in class.
Grade: class participation and homework demonstration (10%), presentation (20%),
project (30%), and exam (40%).
Mini-Project Guidelines
There are some basic steps you need to consider in order to produce a successful
marketing research project. A simple guideline of what you should include in a project is
listed below.
I. You need a managerial question or problem. Statistics is a tool to help you answer the
question. Make sure your topic is related to quantitative marketing and your applied
methodology is related to GLMs.
II. Define your problem or question in clear, specific terms.
III. Develop your hypotheses. What do you expect to happen?
IV. Find out as much as you can about your question.
 Has someone already done work on your question?
 Is your question one for which there is an answer?
 Collect research about your topic.
V. Design your study. Find or develop techniques and measurement instruments that will
provide objective data pertinent to your hypotheses. Decide how you will analyze your
data.
VI. Collect your data. Marketing datasets are widely available online.
VII. Analyze your data following the plan you developed.
 Visualize/summarize data (charts, tables, histograms, and etc.)
 Present your regression models and discuss alternatives
VIII. Interpret your results and draw conclusions relative to your hypotheses based upon
your data and your analysis of the data. What’s the practical implication of your
conclusions for managers?
IX. Write your results—write, rewrite, spell-check, rewrite and . . . PROOFREAD!
Format and Structure for Mini-Projects
1. Your project must be word-processed using a standard font (e.g., TimesNewRoman 12)
and double-spaced. Length is recommended less than twenty pages (including
appendices).
2. All charts, data sheets and graphs should be done on the computer. Graphs and charts
should have a title and properly labeled axes. You need graphs in BOTH your written
project and on your visual aid.
3. When using variables, be sure to state what each represents. Give sufficient
explanation of statistical formula and procedures.
4. All pages must be numbered.
5. Order of pages:
 COVER SHEET—The first page of your project must be a cover sheet that
includes your name(s), date and the title of your project.
 EXECUTIVE SUMMARY—Briefly describe your research topic, conclusions
and practical implications for managers.
 TABLE OF CONTENTS
 INTRODUCTION, PROBLEM, PURPOSE—This section should include the
statement of your question, a statement as to why your project is important or
relevant, and your null and alternative hypotheses. Also discuss what you
expected to find.
 RESEARCH—A summary discussion of your background research and
information on related studies goes here.
 METHODS AND PROCEDURES—This should include data sources and
statistical models you choose.
 RESULTS (ANALYSIS OF DATA)—Your data tables and graphs go here,
interspersed with your narrative. This is also where you include all statistical tests
(including calculations). Graphs and charts should be inserted into the appropriate
location of your paper or referenced by page number if not in the body of the
paper.
 CONCLUSION—This is where you interpret your findings (what can you
conclude or not conclude from your research) and offer managerial insights.
 DISCUSSION— This should include any difficulties that occurred when
collecting or analyzing data, as well as what you would change in a follow-up to
your project. Limitations of your inferences should be discussed. Suggestions for
further work in the area of your topic, as well as recommendations, are mentioned
here.
 References
 APPENDICES—attach anything technical or supplementary.
6. Final project presentation (30 minutes)
7. Hand in a hard copy of your project and send me the electronic version of your datasets,
R codes, the project report and ppt presentation. Grading criteria: make your project
interesting and worth learning!
Presentation
Each team will select two journal papers (see the list in Pages 5-8) and give a formal
lecture based on the papers. In the lecture (30 minutes), you need to explain the research
problem, the motivation, the model, and the solution to the problem. You also need to
compare the paper to existing work and point out the contribution. Feel free to take you
time to clearly explain the paper.
Besides the presenters, the rest of the class will serve as discussants. The main task of a
discussant is to provide a critique of the paper presented: the significance of the problem,
the suitability of the model, the limitations of modeling assumptions, the role that these
assumptions play in obtaining results, and possible extensions. In addition, discussants
should look for common themes or key issues that link related papers and enhance our
understanding of the topic. Lastly, discussants are expected to raise challenging questions
that would guide class discussion.
The presentation will be graded based on how well you motivate the research and how
clearly you explain the model and results.
Topics covered (tentative)
There is no textbook required. The class materials are pooled among miscellaneous
research books and journal papers.
[BMM] Peter Leeflang, Dick Wittink, Michel Wedel, Philippe Naert, Building Models
for Marketing Decisions. Kluwer, 2000.
[QMR] P.H. Franses, R. Paap. Quantitative Models in Marketing Research. Cambridge
University Press, 2010.
Overview of marketing and modeling principles (2 weeks)
•
•
•
•
•
•
•
•
•
•
P. Kotler and G. Armstrong, Principles of Marketing, Prentice Hall, 14E.
Parts I, II and IV of BMM
Little, J.D.C. 1970. Models and managers: The concept of a decision calculus.
Management Science 16 B466-485.
Little, J.D.C. 1975. BRANDAID: A marketing-mix model, Part 1: Structure.
Operations Research 23(4) 628-655.
Little, J.D.C. 1975. BRANDAID: A marketing-mix model, Part 2:
Implementation, calibration, and case study. Operations Research 23(4) 656-673.
Little, J.D.C. 2004. Comments on “Models and managers: The concept of a
decision calculus”. Management Science 50(12) 1854-1860.
Desai, P. and other editors. 2012. Introducing Marketing Science Institute
Research Priorities to Marketing Science. Marketing Science 31(6) 873-877.
Mela, C.F., J. Roos, Y. Deng. 2013. A Keyword History of Marketing Science.
Marketing Science 32(1) 8-18.
Huber, J., W. Kamakura, C.F. Mela. 2014. A topical history of JMR. Journal of
Marketing Research 84(1) 84–91.
American Marketing Association. 2014. Who Went Where? Survey Results.
Historical, philosophical and technical reviews of probability, statistics and induction
logic (3 weeks)
•
•
•
•
•
•
Part III of BMM
J.M. Keynes, A Treatise on Probability: The Connection Between Philosophy and
the History of Science, MACMILLAN, 1921.
F. Black. 1982. Trouble with econometric models. Financial Analysts Journal
38(2) 29-37.
I. Hacking, The Taming of Chance, Cambridge University Press, 1990.
I. Hacking, An Introduction to Probability and Inductive Logic, Cambridge
University Press, 2001.
A. Desrosières, The Politics of Large Numbers: A History of Statistical Reasoning,
Harvard University Press, 1998.
•
•
T.S. Ferguson, A Course in Large Sample Theory, CRC Press, 1996.
Statistic concepts: Sufficiency, strong consistency, asymptotic normality and
Cramer-Rao lower bound of the Maximum-Likelihood Estimate, asymptotic
distribution of the likelihood ratio test statistic, exponential families, mixture
models, EM algorithms, Bayes’ rule, resampling techniques, and empirical
Bayesian estimators
Review of Linear regression models and implementation with R (2 weeks)
•
•
•
•
Part III of BMM
Chapters 1-3 of QMR
G.J. Kerns. Introduction to Probability and Statistics Using R. Online PDF.
Armstrong, J.S. 2012. Illusions in regression analysis. International Journal of
Forecasting 28 689–694.
Empirical research on promotions and pass-through rates (1 week)
•
•
•
•
•
•
Jeuland, A., S. Shugan. 1983. Managing channel profits. Marketing Science 2(3)
239-272.
Tyagi, R.K. 1999. A characterization of retailer response to manufacturer trade
deals. Journal of Marketing! Research XXXVl(November) 510-516.
Besanko, D., J.-P. Dube, S. Gupta. 2005. Own-brand and cross-brand retail passthrough. Marketing Science 24 123-137. (Presenters: )
Nijs, V., Misra, et al. 2010. Channel pass-through of trade promotions.
Marketing Science 29 250-267. (Presenters: )
Ailawadi, K.L., et al. 2009. Communication and promotion decisions in retailing:
A review and directions for future research. Journal of Retailing 85(1) 42–55.
(Presenters: )
Ailawadi, K.L., N. Borin, P.W. Farris. 1995. Market power and performance: A
cross-industry analysis of manufacturers and retailers, Journal of Retailing 71(3)
211-248. (Presenters: )
Logit and multinomial models (2 weeks)
•
•
•
•
•
Chapters 4 and 5 of QMR
Anderson, S.P., A. de Palma, J.-F. Thisse. Discrete Choice Theory of Product
Differentiation. MIT, 1992.
Myerson, R.B. Game Theory. Harvard University Press, 1991.
Thomas, L. Consumer Credit Models: Pricing, Profit, and Portfolios. Oxford,
2009.
Guadagni, P.M., J.D.C. Little. 1983. A Logit model of brand choice calibrated on
scanner data. Marketing Science 2 203-238.
•
•
•
•
•
•
Guadagni, P.M., J.D.C. Little. 2008. Commentary: A Logit model of brand choice
calibrated on scanner data - A 25th Anniversary Perspective. Marketing Science
27(1) 26-28.
J.S. Cramer. 2003. The origins and development of the logit model. Online PDF.
Besanko, D., S. Gupta, D. Jain. 1998. Logit demand estimation under competitive
pricing behavior: An equilibrium framework. Management Science 44 1533-1547.
(Presenters: )
K. Sudhir. 2001. Structural analysis of manufacturer pricing in the presence of a
strategic retailer. Marketing Science 20(3) 244-264. (Presenters: )
Mehta, N., X. Chen, O. Narasimhan. 2010. Examining demand elasticities in
Hanemann’s framework: A theoretical and empirical analysis. Marketing Science
29 422-437. (Presenters: )
Mehta, N., Y. Ma. 2012. A multi-category model of consumers’ purchase
incidence, quantity, and brand choice decisions: Methodological issues and
implications on promotional decisions. Journal of Marketing Research 39 435-451.
(Presenters: )
Bass diffusion models (2 weeks)
•
•
•
•
•
Chapter 8 of QMR
V. Mahajan, E. Muller and Y. Wind (Eds.), New-Product Diffusion Models.
Springer, 2000.
Bass, F.M. 2004 (reprint). A New Product Growth for Model Consumer Durables.
Management Science 50 1825-1840.
Bass, F.M., T.V. Krishnan and D.C. Jain. 1994. Why the Bass model fits without
decision variables. Marketing Science 13 203-223. (Presenters: )
Peers, Y., D. Fok, P. Hans Franses. 2012. Modeling seasonality in new product
diffusion. Marketing Science 31 351-364. (Presenters: )
Hazard rate models (1 week if time permits)
•
•
•
•
Chapter 8 of QMR
Dalal, S.R., J.C. Lee, D.J. Sabavala. 1984. Prediction of individual buying
behavior: A Poisson-Bernoulli model with arbitrary heterogeneity. Marketing
Science 3 352-367.
Chintagunta, P.K. 1993. Investigating purchase incidence, brand choice and
purchase quantity decisions of households. Marketing Science 12 184-209.
Bucklin, R.E., S. Gupta, S. Siddarth. 1998. Determining segmentation in sales
response across consumer purchase behaviors. Journal of Marketing Research 35
189-197.
Project presentation (1 week)
Exam (1 week)