96796 - BUSINESS STATISTICS: METHOD AND APPLICATIONS

Learning outcomes

At the end of the course the student knows the main applied statistical methods to extract useful information from business databases and to support the management decision process in the face of uncertainty. In particular the student is able to apply statistical indicators and linear/non-linear regression analysis for the evaluation of firms’ performance, the study of customers’ behavior and the choice of an investment project, by using a statistical software. At the end of the course the student is able: - to select and apply the most appropriate methodology to quantitatively analyze the business phenomena; - to critically interpret empirical results.

Course contents

By the end of this course, you should be able to: 1. explain what is meant by business statistics, the difference between risk and uncertainty and why they play a fundamental role in the world of business; 2. describe the different kinds of market structure and the implication of market structure for how a business can influence demand for its products/services using pricing, promotion and related strategies; 3. use appropriate data sets to measure and interpret the degree of concentration in a market in which a firm operates and explain how market concentration can affect a firm’s pricing, promotion and related strategies; 4. explain the importance of randomised experiments for making causal inferences and their role in A/B testing for optimising the performance of a website, using frequentist statistical methods; 5. explain what is meant by Bayesian statistics and how randomised experiments can be interpreted from the Bayesian perspective. Explain how the Bayesian approach differs from the frequentist one and use the example of an A/B test to compare and contrast the Bayesian and frequentist approaches; 6. explain the challenges of making causal inferences using non-randomised experimental data sets, such as observational data sets or databases available to your firm and the conditions under which a multiple regression framework can be used to estimate a causal effect; 7. explain the challenges posed to the business statistician by problems arising from endogeneity resulting from omitted variable bias, simultaneity bias and measurement error; 8. explain how natural experiments, instrumental variables and regression discontinuity designs can overcome the endogeneity problem and be able to use appropriate software to carry out instrumental variable estimation; 9. use the above methods to extract and analyse business-relevant information from appropriate data sets and present and report it in a form which provides clear and accessible decision-support to managers. These objectives are arranged under three main headings, each of which contains both taught topics and an application (or applications): 1. Part 1: ‘Know your market’: Market structure and concentration (with applications to measuring and interpreting market structure). 2. Part 2: ‘Know your customers (A)’: Causal effect estimation in randomised experiments (with applications to website optimisation). We take both frequentist and Bayesian approaches to this topic. 3. Part 3: ‘Know your customers (B)’: Causal effect estimation in non-randomised settings (with applications in causal effect estimation for advertising expenditure).

Readings/Bibliography

There is no single textbook for this course.

Introduction to Econometrics, Global Edition. J. H. Stock and M. W. Watson. 4th edition. Pearson International Content, 2019.

Mostly Harmless Econometrics: An Empiricist’s Companion. J. D. Angrist and J.-S. Pischke. Princeton University Press, 2009.

Econometric Analysis of Cross-Sectional and Panel Data. J. M. Wooldridge. MIT Press, 2010.

Principles of Economics. S. A. Greenlaw and D. Shapiro. Openstax, 2011.

Microeconomic Theory: Concepts and Connections. M. E. Wetzstein. 2nd edition. Routledge, New York, 2013.

Introduction to Bayesian Statistics. W. M. Bolstad and J. M. Curran. John Wiley & Sons, Inc., Third edition, 2016

Teaching methods

Lectures and workshops, some computer-based.

Assessment methods

Written examination at the end of the course which will examine students on the learning aims and objectives that are set for the course and each individual class.

Teaching tools

The e-learning platform will provide access to lecture slides, case studies, data sets and solutions. Normally lessons will also be recorded and made available to support your studies. Weekly office hours will be held and comments and grading of the exams will be available in online feedback meetings.

Office hours

See the website of Martin Forster