Decision Analysis
Health/Public Administration 623

Spring 1997
Thursdays 6-8:30 (POT 110)

Instructor:
Dr. Sheila Murray
431 Patterson Office Tower
Phone: 257-5581
E-mail: semurr1@pop.uky.edu
URL: http://www.uky.edu/Classes/PA623

Office Hours:
Tuesday and Thursday 4:30-5:30 p.m.

Philosophy:
My goal is to create together with you an environment that encourages autonomy, personal responsibility, continuous learning and the ability to adapt. There are two principles that guide this goal:

• You and I share responsibility for the quality of your learning
• The motivation, for both of us, should be the satisfaction that derives from improving the quality of your learning.

In order to focus on the quality of your learning and to increase your active participation in the class, we will assess each class period. In the final minute (or two) of each class, I will ask each of you to respond to scale questions and the following two open-ended questions:

• What is the most important thing you learned in today’s class?
• What is the least clear point still remaining at the end of today’s class?

This survey gives me information about what works and provides you with an opportunity to participate in important course management decisions.

Course Description:
This course presumes that students understand the concepts presented in PA/HA 621, especially probability distributions, expected value, variance, use of Excel and familiarity with SAS. Decision Analysis provides a conceptual understanding of the role that quantitative methods play in the decision making process and the economic way of thinking. The course objective is to acquaint you with the basic concepts of the decision making process (such as decision analysis and linear programming) and econometrics (such as regression). The emphasis is non-mathematical and directed toward learning various techniques and their applications. Thus, the concepts will be introduced through problems, cases and projects that use data and computing software (Excel and SAS).

Objectives:

• Understand the basic concepts of the decision making under uncertainty and resource constraints.
• Understand how to use data to make decisions and ask economic questions.
• Learn by doing
  

Skills and Knowledge:

• Determine optimal strategies when there are several alternatives or there is uncertainty.
• Use linear programming to solve management problems
• Use Excel to solve decision problems
• Understand for each econometric problem the nature of the problem, the consequences for
  OLS estimation, how to detect the problem and how to get rid of the problem.
• Understand the questions that can be answered with econometrics
• Build econometric models using dummy variables, alternative functional forms and dummy
  dependent variables
• Use SAS and Excel to estimate regressions
• Write 1-2 page reports that summarize the question, method and results
• Knowledge of sources of data
• Use, manage and document data
  

Texts:
ASW: Anderson, Sweeney and Williams, Quantitative Methods for Business, Sixth Edition (NY: West, 1995)
S: Studenmund, Using Econometrics, Second (or Third) Edition (NY: Harper Collins, 1992)

Course Methodology:
Class time will be devoted to solving problems. There will be a short lecture on the major points of the topic, but the approach to topics is very applied. We only meet once a week, thus it is imperative that you prepare at least two hours for each hour of class. Class is an opportunity for you to ask questions about concepts, techniques and applications. Exams and assignments will include material that is not explicitly reviewed in class. In order for class time to be productive, each student in the class will be assigned at least one problem that he or she will work out before we review the topic in class. These problems will turned in and I will compile them and distribute in the next class period. The problems will be graded (seebelow) and I will call on students to work out the problems in class. We will also meet in the Martin School Computer Lab (Room 402 Patterson Office Tower) in the last half of each class period. In the lab, I will illustrate how to use the statistical software and you will be able to work on problems and projects in class using SAS and Excel.

Course Evaluation:
Your grade will be determined by weighting your performance on two types of writing/analysis
assignments (four manager reports and a data journal), the weekly problems, a midterm exam and a final exam.

Manager Reports:
The reports are case problems that are due throughout the semester. These reports are based on case studies in the text and from handouts that I have prepared. The dates that the reports are assigned are given in the outline that follows. You will have one week to complete the reports, (see late policy). With the exception of Manager Report 4, all data will be given to you. The format of the reports will be detailed on a handout given with each assignment. The first three manager reports each count 7 % of the final grade (21% total). The last managers report is an analysis of an economic question of your choice using data from your data journal. This particular report counts 9% of the final grade.

Problems:
Each student in the class will be assigned at least one problem that he or she will work out before we review the topic in class. These problems will turned in and I will compile them and distribute in the next class period. The problems will be graded on a check-, check, or check+ basis. No problems will be accepted late, I will drop the three lowest scores. The problems count 5% of the final grade.

Data Journal:
The data journal is a log that you will keep throughout the semester. In it you will record your efforts to find, collect, summarize and document data from the SSTARS center, Internet source other than ICPSR and two published sources. The journals will be evaluated twice as listed on the outline. The format of the journal is detailed on a handout. The journal counts 15 percent of your final grade.

Exams:
There are two exams: a midterm that covers the decision analysis material from the first half of the course and a final exam that covers econometrics material from the second half of the course. Both exams will have a skills section in which you will solve problems similar to the end of the chapter problems and a report section in which you will analyze a problem and write a short version of a managers report. Each exam counts 25 % of your final grade.

Late Policy:
Assignments are due at the beginning of the class period. I will accept manager reports and the data
journal up to 1 week after they are due, however, a 10 % penalty will be assessed to the grade. Reports more that one week late will not be considered for a grade. Makeup exams will be given only for university defined excused absences.

Scale:
90-100 ...A
80-89 ...B
70-79 ...C
60-69 ...D
0-59 ...E

Class Outline and Schedule of Assignments:

Jan. 16
Introduction and Course Outline
Review of 621 Skills
Readings: ASW: Chapter 1-3, Excel Handout

Jan. 23
Decision Analysis
Readings: ASW: Chapter 4;
Computer Application: Decision Analysis and Spread Sheets, ASW pp. 147-50

Jan. 30
Decision Analysis
Readings: ASW: Chapter 5
Computer Application: Example of Data Collection from ICPSR, Individual Work on Managers Report 1

Feb. 6
Linear Programming
Readings: ASW Chapters 7 and 8
Computer Application: Spreadsheet Solution of Linear Programs, ASW pp. 335-38
Managers Report 1 due

Feb. 13
Linear Programming
Reading: ASW Chapter 9
Computer Application: Spreadsheet Solution of Linear Programs, ASW pp. 391-393, Work on Managers Report 2

Feb. 20
Overview of Regression
Reading: S Chapters 1 and 2, Handout on SAS
Computer Application: Using SAS
Managers Report 2 due

Feb. 27
Midterm

Mar. 6
No Class

Mar. 13
Learning to Use Regression
Reading: S Chapters 3-5
Computer Application: Estimation and Hypothesis Testing Using SAS
Data Journal 2 Published Sources and Proposals Due

Mar. 17-21
Spring Break

Mar. 27
Specifying a Regression Model
Reading: S Chapter 6
Computer Application: Obtaining Summary Statistics Using SAS, Predicting Physician Utilization

Apr. 3
Specifying Functional Form
Reading: S Chapter 7
Computer Application: Investigating Wage Differentials

Apr. 10
Forecasting
Reading: ASW Chapter 6, S Chapter 15, Section 1
Computer Application: Work on Managers Report 3
Data Journal Due

Apr. 17
Dummy Dependent Variables
Readings: S Chapter 13
Computer Application: Computing Marginal Effects
Managers Report 3 due

Apr. 24
Efficient Estimation
Readings: S: Chapter 8-10
Computer Application: Testing for Heteroskedasticity and Serial Correlation with SAS

May 1
Review of Regression
Readings: S: Chapter 11
Managers Report 4 due

May 8
Final Exam