Public Affairs 7571: Multivariate Regression Analysis

Back to All Courses

This is a sample syllabus to provide general information about the course and it's requirements. Course requirements are subject to change. This syllabus does not contain all assignment or course detail and currently enrolled students should reference the syllabus provided by their instructor. For a specific syllabus, please email us a request.

Back to All Courses

Course Overview

3 Credit Hours
Modalities Available: In Person

The course is for students who are interested in a career in government or the non-profit sector who
need more advanced statistical skills or for students interested in research or academia who need to
build a strong foundation in applied regression analysis. It provides students with knowledge about the Ordinary Least Squares model and introduces students to a wide range of additional models that can be applied to typical policy or management applications. The course is particularly useful for master’s students who will write a capstone research paper or master’s thesis and doctoral students who plan to conduct applied research.

The primary objective of this course is to provide an understanding of regression techniques, both from a “producer’s” and a “consumer’s” perspective. Regression, which is a statistical method used to study the relationships among two or more variables, is the most widely used statistical technique in public policy analysis. It is almost impossible to learn statistical concepts without lots of practice using statistical methods, so the course focuses on applying the theory through multiple hands-on exercises, both inside of class in a lab format and outside of class with homework assignments. These exercises will also give students practice in how to use data from external sources and report statistical results in a clear manner. By the end of the course, students should not only be intelligent consumers who can readily interpret regression analysis performed by others, but they will also be equipped to test research hypotheses involving relationships among multiple variables and to relay their results to relevant stakeholders. Students should also be able to able to identify the appropriate models to use based on the nature of their data.

Learning Outcomes

Upon completion of the course, students should understand

Basic assumptions of the Ordinary Least Squares Model
Violations of the assumptions of the Ordinary Least Squares Model
Regression models with continuous, categorical, count, or limited dependent variables

Upon completion of the course, students should be able to

Apply techniques of regression analysis to test research hypotheses
Effectively communicate the results of conducting regression analysis
Use the appropriate regression model given a particular dependent variable

Requirements and Expectations

Required Materials:
A textbook may be required for this course. Consult your instructor's syllabus for details.

Mode of Delivery:
This is an in-person course that meets twice weekly in the classroom on Mondays and Wednesdays. In a typical week, Mondays will be lectures and Wednesdays will the completion of the week’s lecture and a hand’s on lab applying that week’s material. Except for textbook readings, and the exams, almost all course content and materials are accessible and administered through the CarmenCanvas course website.

Pace of Activities:
The course is divided into weekly modules. Each weekly module contains the topics associated with the two class meetings for that week. The weekly modules and class meeting topics are specified in the Course Schedule below and the CarmenCanvas course website. Students must complete readings and other course materials assigned for a class meeting prior to the meeting time.

Credit Hours and Work Expectations: This is a 3 credit-hour course. For each credit hour, there should be about an hour of in class meeting time and 2 hours out of class work. For a three-credit class, you can expect an average of 6 hours of work outside of class completing readings, assignments and preparing for exams. This totals 9 hours per week for an average grade of C.

Statistics Review:
This class assumes familiarity with basic statistics. Key relevant statistical concepts will be reviewed during week two.

Homework assignments (best 6 of 7), 150 points (25/each), 25%
Midterm Exam 1, 100 points, 25%
Midterm Exam 2, 100 points, 25%
Research paper, 100 points, 25%
Total Points: 450

Attendance
Description: Students are expected to attend each class. While you do not receive an explicit grade for class participation, in my past courses, I have found there to be a strong correlation between class participation and grades in the other course components. Further, if attendance becomes a problem (i.e. you begin chronically missing class), I reserve the right to deduct up to 25 points off your final grade.

Homework
Description: The homework grade will be based upon the six highest homework grade scores. While students may collaborate on homework assignments, each student must turn in a separate assignment with his or her own answers. Assignments are due at the beginning of class, and late assignments will not be accepted. For your homework assignments,

Assignments will be posted and turned in on CarmenCanvas
To receive full credit, show all work
Feel free to use Stata as much as you can/want to
When you use Stata to answer a problem
- Provide the Stata commands and output as part of what you turn in (screen shots work best)
  - This is “showing your work” for Stata problems
  - Minimize the amount of output (and number of pages) if at all possible
- Make sure that you also directly answer the question
  - For example, it is not enough to provide the Stata output that shows that a mean is 1234. You should also tell me that the mean is 1234, as is indicated in the Stata output.

Exams
Description: The two mid-term exams will be in-class. Students are allowed to use one sheet of notes (more direction will be given in class). Collaboration on the exam is strictly forbidden (and unnecessary given the open notes policy). Make up exams will be given only if arrangements are made in advance.

Research paper
Description: The paper is intended to help integrate the course material and provide students an opportunity to demonstrate that they can set up a testable research hypothesis, test the hypothesis, and correctly interpret the results. Students should also demonstrate an awareness of the limitations of their analysis. Detailed instructions for the research paper will be distributed separately and posted to CarmenCanvas. An intermediate deliverable for the paper will be integrated into one of the weekly homework assignments.

Labs
Description: The hands-on labs are designed to help you practice using the concepts from each class. They are also useful for completing assignments and for preparing for exams. They are ungraded, but I reserve the right to start grading them without prior notice if I find students are not participating.