School of Policy, Planning, and Development

University of Southern California

PLUS 502

Statistics and Arguing from Data

Fall 2002

Course Description

	Class	Laboratory
Days	Tuesdays & Thursdays	Fridays
Times	4:00 – 5:50pm	1:00 – 3:00pm

	Principal instructor	Teaching Assistant
Name	Eric J. HEIKKILA	Lanlan WANG
Position	Associate Professor	Ph.D. Student
E-mail	heikkila@usc.edu	lanlanwa@usc.edu
Website	http://www-rcf.usc.edu/~heikkila/	http://www-rcf.usc.edu/~lanlanwa/
Telephone	1-213-821-1037 (or x11037)

Purpose:

This course teaches you to explore planning issues empirically and to reason with data about such issues. A primary focus is on statistical methods, including regression analysis, for analyzing variance and covariance. A related purpose is to sharpen your ability to evaluate empirical studies and reports undertaken by others. Although many of you may not choose to make statistical analysis your field of specialization, you will very likely be required regularly in your professional career to review empirical studies or reports as inputs to your own decision making. This course gives you a proper foundation for doing so.

Objectives:

By the end of the semester you should be able to:

• Access, download, and organize data sets from the Census and other sources
• Demonstrate a robust understanding of the fundamentals of statistical inference
• Articulate arguments about planning issues in empirical terms and translate those arguments into testable hypotheses
• Work competently with Excel, SPSS for Windows, and other software to perform routine statistical analyses designed to test such hypotheses
• Interpret the output of such statistical analyses in the context of the planning issues that motivated them
• Summarize, interpret, and critique constructively planning reports and articles that employ statistical methods

Required Texts and Readings:

One text book has been ordered for purchase in the University Bookstore:

Sam Kash Kachigan, 1986, Statistical Analysis: An Interdisciplinary Introduction to Univariate & Multivariate Methods, New York: Radius Press.

Any additional assigned readings will be made available to you in a course reader format.

Course Format:

The class meets twice a week for lecture presentations and once a week for laboratory teaching assistance. All three weekly meetings are mandatory. The lectures focus primarily on a presentation of fundamental theoretical concepts in statistics and of regression analysis and other statistical methods used routinely in the empirical analysis of planning issues. The lab meetings will help you to quickly begin working with data sets of your own and assist you with steady progression towards completion of your term project.

Course requirements:

The best way to learn statistics is through active engagement with the material. You will be asked (required) to do so in several ways:

1. Participation – Attendance will be recorded promptly at the beginning of each class, and I except you to attend regularly and punctually. Your are expected to contribute to class discussions in a manner that enhances the learning experience for your fellow students, and so your in-class contributions will be judged on the basis of quality and quantity.
2. Assignments – You have five assignments in all. Three assignments early in the semester will help you learn, respectively, to (i) access and download data sets from the Census Bureau, (ii) use spreadsheets to calculate co-variance and related measures for selected variables from your data set, and (iii) use SPSS for Windows to perform regression analyses with your data set. Two additional assignments following the midterm exam will direct you, respectively to (iv) provide an extended abstract of your proposed term project and (v) write an appropriate literature review relevant for your term project.
3. Midterm exam – The midterm exam will test your fundamental knowledge of regression analysis and related concepts as discussed in class during the first part of the course.
4. Term project – Your term project is in the form of a professional planning report that employs at least one of the statistical methods discussed in class, using a planning topic and data of your own choosing. You are responsible for (i) selecting a suitable topic, (ii) formulating an empirical argument, (iii) locating and obtaining a relevant data set, (iv) utilizing relevant statistical methods, and (v) interpreting the results of your analysis in the context of the planning topic you have identified. While this project is due near the end of the semester, you should begin planning your term project early on.
5. In-class presentations – Each of you will be asked to make an in-class presentation of your term project near the end of the semester.

Grading:

Your course grade will be calculated as follows:

1	Participation (attendance 5%; discussion 5%)	10%
2	Assignments (five @ 5%)	25%
3	Midterm exam	15%
4a	Term project – draft	15%
5	In-class presentations	15%
4b	Term project – final	20%
	Total	100%

Course outline and corresponding readings:

· Data sets – Kachigan, chapters 1 through 4

· Variance and z-scores – Kachigan, chapter 5

· Correlation and co-variance – Kachigan, chapter 5 and chapter 10

· Central limit theorem – Kachigan, chapter 6

· Analysis of variance – Kachigan, chapter 12

· Regression analysis – Kachigan, chapter 11

· Factor analysis – Kachigan, chapter 15

· Discriminant analysis – Kachigan, chapter 14

· Cluster analysis – Kachigan, chapter 16

· Maximum likelihood – Kennedy[1], chapter 2

· Discrete choice models – Kennedy, chapter 15

You are expected to read the assigned readings before each class[2], and your in-class discussion should make it evident that you have done so.

Class schedule:

Please refer to the “Schedule of Topics” that I have prepared as a separate attachment to this syllabus.

---

[1] Peter Kennedy, 1992, A Guide to Econometrics, MIT Press, 3^rd edition, Cambridge, MA. To be included in course reader.

[2] I make an exception for the first class – but only for the first class!