OR/MS Today - February 2000 - Issues in Education

February 2000

Executive Education Opportunities in Management Science

By Wayne L. Winston

In many business schools, management science education has thrived in recent years. Teaching more modeling and problem-solving ideas within the context of spreadsheets and spreadsheet add-ins seems to have revitalized the field. I believe this creates an opportunity that most of us miss.

Virtually no consultant financial analyst, marketing analyst or operations analyst who graduated more than five years ago knows much about the incredible opportunities for problem solving that exist on their notebook or desktop computer. There is an opportunity for OR/MS and statistics professors to "reach out and touch" the corporate world. Companies want their analysts to be more efficient and aid in making better corporate decisions. Yet most analysts know nothing about optimization, data analysis using tools such as pivot tables, simulation and decision-making under uncertainty. These tools are critical to making better business decisions in areas such as launching new products, optimizing distribution or capacity planning.

I have been fortunate in being well paid for conducting "missionary training" (otherwise known as executive education) at several U.S. companies. In this column I would like to discuss some of the courses I have developed and some opportunities for future course development.

Eli Lilly Two-Day Course in Spreadsheet Modeling: During the past five years I have taught a two-day course in spreadsheet modeling at Eli Lilly, one of the 10 largest drug companies in the world. My audience was primarily scientists, engineers and MBA graduates. A summary of the curriculum follows:

Six hours of spreadsheet simulation models including the newsperson problem, project management, new product valuation, Deming's funnel experiment, and scheduling production to meet demand on time.
Six hours of optimization models including product mix, aggregate planning, manpower scheduling, optimization of process yield, portfolio optimization, facility location and rating sports teams.
One hour introduction to genetic algorithms and neural networks.
Three hours of EXCEL "tips" including Data Tables, Pivot Tables, Lookup functions, Goal Seek and Trend Curves.

Participants have developed models to schedule production of drugs within plants and on a global basis.

Monte Carlo Simulation at General Motors: Under the leadership of President Rick Waggoner, General Motors is attempting to use Monte Carlo Simulation (with @RISK) as an aid in decision-making under uncertainty. For example, GM runs a Monte Carlo simulation to evaluate the risk and expected return associated with each new car model. In our one-day introduction to Monte Carlo simulation we cover the following topics:

Introduction to Monte Carlo simulation with the newsperson problem. We use this example to introduce confidence intervals and a sampling distribution and we use discrete and normal random variables.
Simulating profitability of a new product. This example introduces cumulative graphs, tornado graphs, probability of meeting hurdle rate, and analysis of simulation statistics using triangular and normal random variables.
Modeling uncertainty with general distributions (risk general and risk cumulative).
Modeling correlations.
Forecasting structural costs.
Teaching Spreadsheet Modeling at Microsoft

This summer I have been busy developing a spreadsheet-modeling curriculum for Microsoft's financial analysts. This has been a great experience. Here is how the curriculum looks so far :

Course I: Data Analysis with EXCEL (six hours)

Histograms and Descriptive Statistics
Curve Fitting (linear, power, exponential and moving average) with the trend curve.
Multiple Regression
One Way ANOVA
Pivot Tables

I am convinced that Pivot Tables are the single most useful feature in EXCEL, yet few business analysts know they exist.

Course II: Basic Business Modeling (6 hours):

Data Tables, Auditing Tool and Goal Seek.
Handling circular spreadsheets.
The Scenario Manager.
Extensive coverage of the EXCEL Solver.
Coverage of neat EXCEL functions such as MATCH, INDEX, SUMIF and COUNTIF.
Introduction to Monte Carlo simulation (introduced as a generalization of the Scenario Manager).

We are still designing an advanced business modeling class. Modules involving advanced coverage of Solver, advanced Monte Carlo analysis and real options are under consideration. To my knowledge this is the first attempt by a major U.S. company to develop a coherent curriculum that will keep their analysts' skills current.

Wayne L. Winston is a professor of Decision Sciences at Indiana University School of Business, Bloomington, Ind.

This is a regular column sponsored by INFORMED, the INFORMS Forum on Education. Contact Armann Ingolfsson at armann.ingolfsson@ualberta.ca

Table of Contents

OR/MS Today Home Page

Lionheart Publishing, Inc.
506 Roswell Street, Suite 220, Marietta, GA 30060, USA
Phone: 770-431-0867 | Fax: 770-432-6969
E-mail: lpi@lionhrtpub.com
URL: http://www.lionhrtpub.com