York University
Schulich School of Business

Winter 1999 --- Prof. Peter Tryfos


The objective of MGTS 5120 is to provide an introduction to quantitative methods for business research, analysis, forecasting, and optimization. By "research" we refer to the gathering of information necessary for making business decisions, often by means of samples and surveys. By "analysis" we understand the process of reducing the usually large quantity of information so gathered to manageable measures and patterns. "Forecasting" refers to the extrapolation or projection of these measures and patterns. Finally, by "optimization" we understand the selection of the best of the available alternatives. Research, analysis, forecasting, and optimization are important elements of any business problem. Of course, MGTS 5120 cannot cover all methods, nor does it address all types of business problems---only problems which can benefit from the application of quantitative (mathematical, statistical) methods. The computer is an indispensable tool for the implementation of these methods.

The course aims at providing not only a good understanding of basic methods, but also an appreciation of some of their applications to business problems. In general, applications will take the form of readings, cases, or computer simulations. Applications constitute a key element of the course. Students are tested for their understanding of both the methods and the applications to which they were exposed in the course.

The following are distributed as course materials:

(PT1) Peter Tryfos, Lecture Notes. [Chapter 1: Introduction, and Chapter 2: Preliminaries (yellow pages) are adapted from P. Tryfos, Sampling for Applied Research: Text and Cases, Wiley, 1996. Chapter 1: Summarizing data, and Chapter 2: Probability and Probability Distributions (white pages) are adapted from P. Tryfos, Business Statistics, McGraw-Hill Ryerson, 1989. Chapter 3, Regression (white pages) is adapted from P. Tryfos, Methods for Business Analysis and Forecasting: Text and Cases, Wiley, 1998.]

(GS) Gordon C. Shaw, The Meaning of Linear Programming.

(PT2) Peter Tryfos, Superior Insurance Inc. (a computer-based case/simulation with associated programs and instructions).


I. Sampling

Sources of business information; populations, samples and surveys. Random vs. non-random sampling. Some types of samples: simple, stratified, two-stage. Considerations for selecting a sample design. Computing.
(PT1: Chapters 1 and 2, yellow pages)

II. Data Analysis

The need for data reduction. Variables and attributes. Univariate frequency and relative frequency distributions. Measures of location and dispersion. Graphics. Bivariate distributions and cross-tabulations. Independence, correlation. Scatter diagrams. Probability and probability distributions. Computing.
(PT1: Chapters 1 and 2, white pages)

III. Measurement of Relationships

Simple and multiple linear regression. Measures of fit and contribution. Attributes as explanatory variables---dummy variables. Basic time series analysis for trend and seasonality. Computing.
(PT1: Chapter 3, white pages)

IV. Optimization

Introduction to optimization methods. Linear programming. Computing.

Students are free to use whatever software package they prefer. However, support will be provided only for Excel for Windows and its accompanying statistical add-on.


In the Fall 1998 term, all assignments are related to the Superior Insurance case. See attached schedule.


Students may work in teams of one to four members. Students are free to select the members of their team. Each team will write one report and prepare a 10-minute presentation. A student’s mark will be calculated as follows:

Item Weight (%)

Team report: 80
Team performance: 10
Team presentation: 5
Class participation: 5
Total 100

The team performance mark is based on the profit achieved in the Superior Insurance case in relation to that of other teams.

The team mark may be distributed equally or differentially among team members (see attached instructions).

The class participation mark is based on the instructor's assessment of the quality of the student's individual contribution in class discussions.


Prof. Peter Tryfos
Rm. 335 SSB
Tel. (and voice messages): (416) 736-2100 x 77949
E-mail: ptryfos@bus.yorku.ca
Web home page: http://www.bus.yorku.ca/faculty/tryfos/hpage.htm

Secretary: Ms. Paula Ironi
Rm. 344 SSB
Tel.: (416) 736-5074


All assignments are due at, or must be completed by, the beginning of class on the dates indicated below. Marked with a ** are assignments that require written submissions, execution of a computer program, or presentation. Marked with a ^ are questions intended for that week’s class discussion. All assignments are subject to change with notice.

SI # = Superior Insurance assignment number (see text of case)
SDF = Summary of Decisions Form (Superior Insurance Appendix C)

Due Week
Jan. 18
* Read Chapters 1 and 2 (Yellow pages).
* Read case Superior Insurance Inc.
** Form teams, submit Team Form (team members, password, name).
^ How many IAB records should be purchased?
Jan. 25
* Read Chapters 1 and 2 (White pages).
* Prepare draft of SI #1
** Determine the size of sample of IAB records and run program SUPINS0.
^ To which driver and policy characteristics do the IAB data shed light?
Feb. 1
* Read Chapter 3 (White pages).
* Prepare draft of SI # 2.
^ How would you forecast the loss of a driver with given characteristics?
Feb. 8
* Read The Meaning of Linear Programming (by G. C. Shaw, white pages).
* Prepare draft of SI # 3.
^How should the premium formula be set under the conditions of SI # 4 and SI # 5?
Feb. 12
** Submit SDF (SI Appendix C) for SI # 4 and 5
Feb. 15
**Submit written report to the management of Superior Insurance covering SI # 1 to 5 and describing the methods applied, alternatives considered, the rationale for decisions made, and the expected results in 1999.
** Prepare a 10-minute presentation of the highlights of the written report.