ECO 310 Econometrics
Spring 2013 
Dr.
Robert Jantzen
Economics Department 
Where and WhenIn the Spring of 2013 this course meets at 9:30 a.m. on Mondays and Wednesdays in Amend 107. Classes begin on 1/16/13.A laboratory approach to multivariate research, with an emphasis on economic applications. A review of the basic concepts of regression analysis and the testing of hypotheses, and the statistical problems that arise when the simple regression model's assumptions are violated by the data or the model being analyzed, and appropriate countermeasures. An examination not only of the classical regression model, but also qualitative choice and simultaneous equations models. Instruction in the basics of computerized data analysis, including collection, coding and statistical programming. PREREQUISITE: any onesemester Introduction to Statistics course. 3 credits. Course ObjectivesThe primary objective of this course is to impart to students a working knowledge of how best to analyze simple and multivariate relationships using a variety of regression models. Students will learn how to test and correct for a wide variety of standard statistical problems that appear when data is analyzed. These include serial correlation, multicollinearity, heteroskedasticity, specification bias, measurement error, among others. Students will not only master the statistical theory, but also the computer programming skills necessary to fieldtest multivariate models.This course will rely principally on lecture and discusssion, augmented by statistical programming demonstrations and labs. Texts:
Jantzen, R. A Brief Guide to the Gretl Program. Unpublished manuscript, 2012. Jantzen, R. A Brief Introduction to Multiple Regression. Unpublished manuscript, 2004. Jantzen, R. A Brief Guide to Classical Regression "Problems". Unpublished manuscript, 2004. Additional required readings will be assigned by the instructor at appropriate times in the course. Course Requirements and Grading:Student grades in this course will reflect assessment in the following areas: Homeworks (relative
weight = .2)
All students must complete the term project, including
a short oral presentation during the last week of classes. The lowest
grade on one of the above 5 areas, except for the term project, will be
Academic dishonesty will be penalized heavily. Plagiarism (the copying of text from other sources without the use of quotation marks) and/or cheating will result in a grade of F for the paper/exam involved. In addition, students having excessive absences (6 or more) will receive the grade of FA (failed for absence). Being late to a class will count as an absence. I. Description The fundamental purpose of the term paper is for the student to utilize multiple regression analysis to assess the relationships between a dependent variable and at least three explanatory variables. The term project must be written in the student's own words, be typed (double spaced) and contain an appendix that includes all of the statistical outputs utilized to generate the tables and tests described in the paper. Term projects must be submitted to the instructor via email as a single MSWord or Adobe PDF file. In addition, each student must make an oral presentation of the term project to the class during the last week of classes. II. Organization The term paper for this course must contain: A. an Introduction that briefly explains the purpose of the paper. B. a Review of the Literature section that reviews the methods and findings of at least one other study that has already examined the topic of your study. C. a Data and Methodology section that explains the sources of the data, their time and scope, and the model to be estimated. This section must also detail the expected relationships between the variables, and expound on any anticipated statistical problems and their appropriate corrections. D. an Empirical Results section that provides and discusses: i. descriptive statistics concerning the model's variables. ii. an analysis of tests for heteroskedasticity or serial correlation, and corrections, if appropriate. iii. F or Likelihood ratio tests on the overall model iv. overall goodness of fit statistics. iv. T tests for each population regression coefficient. v. estimated coefficients and their confidence intervals, using the appropriate regression results. v. standardized coefficients. vi. an analysis of the likely effects of specification bias on the coefficients of the estimated model. Specifically, identify one plausible explainer that was not included in the estimated model and how its exclusion would affect the coefficients that were estimated. E. a Summary and Conclusions section that highlights the key findings and policy implications of the study, if any. Click on the following link for an example of a sample project: sampleproject.doc


III. Data: To satisfy the term project requirement, students must collect and analyze their own data. Click here for information on how to search, download and organize data from the web.

(approximate) 



1/16 & 1/21  Introduction  Chapter 1.  Homework 1 and Homework 2 

Multiple Regression  Chapters 2 and 3.  Homework 3 and Homework 4 

Hypothesis Testing  Chapter 5.  Homework 6 , Homework 7 and Homework 8 

Exam#1  

Dummy Variables and Nonlinear Models  Chapter 7.  Homework 9 and Homework 10 

Regression Assumptions and Estimator Properties  Chapter 4.  Homework 5 

Specification Bias  Chapter 6.  Homework 11 

Multicollinearity  Chapter 8.  Homework 12 

Exam#2  

Heteroskedasticity & Normal Residuals  Chapter 10.  Homework 13 

Serial Correlation  Chapter 9.  Homework 14 

Qualitative Choice Models  Chapter 13.  Homework 15 

Panel Data Models  Chapter 16.  Homework 16 

Forecasting Trend Models  Chapter 15 and Handout  Homework 17 

Exam#3 
Software and Data: The Gretl econometric program will be the "platform" used by this course to analyze data. The program is a freeware opensourced program that performs a wide variety of statistical tests that economists utilize. Information on how to download, install and operate the program can be found by clicking here. Assigned homeworks will contain links to the Excel data sets needed for completion. Bear in mind that the Gretl program can only process Excel data sets that have been saved as MS Excel Comma Delimited File (csv) Worksheets.
