Скачать 50.93 Kb.

School of Policy, Planning and Development University of Southern California PPD 558—Multivariate Statistical AnalysisSpring 2011Instructor Contact InformationHalil Toros, Ph.D. Tu 6:00 – 9:20 pm, WPH B36 Office hours: Tu 5:006:00 pm and 9:3010:30 pm at the lab area Email: toros@usc.edu TA: Weijie Wang Email: weijiewa@usc.edu Course Description and ObjectivesThe course begins with a brief review of some of the basic statistical analysis techniques and tools used to describe and interpret data. The primary focus of this course will be on the use of regression analysis, a statistical technique for quantifying and making inferences about relationships between variables. Regression analysis has been increasingly used to evaluate and provide quantitative assessments of existing public policies and programs. In the second half of the semester we will advance to more complex methods of statistical analysis, as you will inevitably find that the complexities of human and institutional activities frequently demand a more sophisticated strategy for the production and analysis of information that can inform public policy decisionmaking and our understanding of policy impacts. The amount of data collected from or about individuals, communities, and institutions has increased rapidly in recent years, and these data are increasingly being used to make major decisions regarding people's health, education, employment, environment and other aspects of social welfare. The principal objective of this course is to train students to be competent and thoughtful users of applied econometrics in support of effective program evaluation, policy analysis and decisionmaking. The course will make students familiar with basic econometric methods and will also introduce some advanced techniques to be applied in public policy research. In fifteen weeks, students will not become an expert in all of the statistical methodologies we will study. However, it is expected that they develop a good understanding of how to design and undertake an empirical analysis in response to a policy or research question and a respectable level of skill in the application of multiple regression analysis. The orientation of the course is applied. Hence, we will be mainly concerned with using, rather than proving, the theory. Students will know which estimation technique is appropriate for a specific question and understand how to verify the accuracy and robustness of estimations using these techniques. In addition, students will take on the role of a researcher or policy analyst and complete an empirical analysis by applying techniques presented in this course to a real problem in a project that will span the duration of the course. This will be a group project (each group will be composed of two students). Students are required to select a problem/decision of interest to them and formulate the problem so that it can be addressed through regression analysis. In this group project, students will formulate one or more policy or research questions to address, explore their data (using descriptive statistics), specify hypotheses to test research questions empirically, identify statistical methods appropriate for their data and analysis, specify statistical models to test, interpret the results of their statistical analyses in terms of the research questions and hypotheses defined at the onset of the study and finally make a power point presentation in class of their study findings, including a discussion of their analytical approach. Finally, students will be assigned some readings with the intention of increasing their understanding of the materials they are studying and put them in the context of actual research. It is one of the objectives of the course to enable students read, understand and interpret academic articles in applied quantitative and regression analysis. To undertake this work, it will be essential for you to become familiar in using two statistical application software packages—SAS & STATA. You will have considerable support for your statistical programming/processing activities via sample programs, supporting documentation, inclass demonstrations, and handson lab sessions. In addition, in the second week of class, the section will be used for SAS & STATA training session. Students may also send emails to the instructor and/or the TA anytime regarding any question on how to use SAS or STATA. It will be up to students to take advantage of these resources for their project. Any student requesting academic accommodations based on a disability is required to register with Disability Services and Programs (DSP) each semester (http://www.usc.edu/studentaffairs/asn/DSP/index.html). A letter of verification for approved accommodations can be obtained from DSP. Please be sure the letter is delivered to me (or to TA) as early in the semester as possible. DSP is located in STU 301 and is open early 8:30 a.m. – 5:00 p.m., Monday through Friday. The phone number for DSP is (213) 7400776. Course RequirementsInitial questions: After the first class, please submit a short note or CV (by email) outlining your education and job history, your background in statistical analysis, and your objectives in taking this course. This information will be used to group students for their projects. If there are students to form their own group please let me know. Readings: The required text for the course is Using Econometrics, Sixth Edition, by A.H. Studenmund, which is available at the bookstore. In addition a few articles will be assigned during the semester. There will be reading assignments with some questions on two of these articles. All lecture notes, articles, assignments and datasets will be available via the blackboard site of USC at http://blackboard.usc.edu. Statistical Software: The course will be taught using SAS and STATA software packages. The students are free to use any of the two software packages for their assignments and project. The instructor will run primarily SAS applications in the class and will show handson how to apply econometric methods using SAS. There will be also sessions and examples using STATA. All classes will be held at the computer lab (Wayne Phillips Hall B36) so that students will be able to learn how to run SAS and STATA programs during the class time and they will solve several exercises using the software. The lab PCs has installed copies of both packages. If students wish to acquire their own copies, SAS is freely available through USC and STATA is available for $98/year. You may use the below link to purchase these software packages.
Grades: The course grade will be based on several components: Inclass exercises, problem sets, reading assignments, a final exam and a group project. Class Participation and InClass Exercises (20%) Students are encouraged to participate in class discussions and ask questions. There will be inclass exercises or quizzes on the end of various sessions to verify that students understood the topics presented that lecture. All inclass exercises and quizzes will be open book and open note. Problem Sets (20%) Students will be assigned four problem sets (5% each), which will be due in two weeks. Problem sets are shown in the weekly course schedule below. Reading Assignments (10%) Students will be assigned two reading assignments (5% each) with some questions to be answered based on the articles. Reading assignments are shown in the weekly course schedule below. Exams (20%) There will be one final exam which will be takehome. Group Project and Presentation (5%presentation & 25%report) As noted earlier, students will complete an empirical analysis in a group project. Each group will be composed of three or four students. The report will include an executive summary, a background section, a literature review, a description of the data, a results section, a conclusion section with policy implications and technical appendices. Students are required to provide constructive feedback (anonymously) to all oral presentations. These feedbacks will be part of the grade and will be due at the end of class on each day of presentations. Due dates of this project are presented in the weekly course schedule below. Each presentation is expected to be approximately half an hour and there will be Q & A session following each presentation. Since data collection may prove to be time consuming students should start early. One useful site is www.stats.org, which has a set of links to major sources of data and statistics. It also includes a number of interesting articles critiquing the ways statistics are abused in the policy process and the media. There are various other sites that may be useful for students to collect data including Census Bureau. Course ScheduleWeek 1—January 10 Review of syllabus, introduction to econometrics and review of statistical methods Lecture Notes Week 2—January 17 LAB session on SAS & STATA Software Lecture Notes Week 3—January 24 Exploratory Data Analysis and Introduction to the Regression Analysis Studenmund, Ch. 12, Lecture Notes [Self Review Ch. 3] Group formation for the class project is dueProblem Set 1 assigned (on Simple & Multiple Regression) Week 4— January 31 Multiple regression Analysis (Continued), Violation of Assumptions of Regression Model & Regression Diagnostics (I)—the Classical Assumptions, Residual Analysis, Problem of Outliers Studenmund, Ch. 4, Lecture Notes Week 5—February 7 Violation of Assumptions of Regression Model & Regression Diagnostics (II) — Multicollinearity, Heteroskedasticity, Checking Normality Studenmund, Ch. 8 & 10, Lecture Notes Problem Set 1 is due Problem Set 2 is assigned (on Regression Diagnostics) Week 6—February 14 Model specification and functional form Studenmund, Ch. 67, Lecture Notes Group project description is due—one page description of the proposed analysis that will include research questions and data to be used Reading Assignment 1 is assigned Week 7—February 21 Model specification and functional form (cont.) Studenmund, Ch. 67, Lecture Notes Problem Set 2 is dueProblem Set 3 is assigned (on Model Specification)Week 8— February 28 Interactions & Dummy Variables, Categorical Data Analysis & Associations Lecture Notes Reading Assignment 1 is due Week 9—March 6 Logistic Regression Studenmund, Ch. 13 & Lecture Notes Problem Set 4 is assigned (on Logistic Regression)Problem Set 3 is dueReading Assignment 2 is assigned Week 10—March 13 SPRING RECESS Week 11—March 20 Time series and longitudinal data analysis, pooling cross sections across time and serial correlation problem Studenmund, Ch. 9, Lecture Notes Problem Set 4 is dueProgress report on class project is due Week 12—March 27 Dynamic, Randomized Experiments and Mixed Models Studenmund, Ch. 16 & Lecture Notes Week 13—April 3 Special Topics in Regression Analysis Lecture Notes Week 14—April 10 Project presentations – I Final Exam is assigned Week 15—April 17 Project presentations – II Reading Assignment 2 is due Week 16—April 24 Lab Session for project help, there will be no lecture Each student will email a grade (A, A or B) for each presentation. May 8 (midnight) Final Exam is dueFinal Project is due ACADEMIC RESPONSIBILITY"Students, faculty, and administrative officials at the University of Southern California, as members of the academic community fulfill a purpose and a responsibility. The University must, therefore, provide an optimal learning environment, and all members of the University community have a responsibility to provide and maintain an atmosphere of free inquiry and expression. The relationship of the individual to this community involves these principles: Each member has an obligation to respect: 1. THE FUNDAMENTAL HUMAN RIGHTS OF OTHERS 2. THE RIGHTS OF OTHERS BASED UPON THE NATURE OF THE EDUCATIONAL PROCESS 3. THE RIGHTS OF THE INSTITUTION ACADEMIC DISHONESTY The following statements and examples explain specific acts of academic dishonesty.
a. Communicating in any way with another student during the examination. b. Copying material from another student's exam. c. Using unauthorized notes, calculators or other devices. 2. Fabrication: Any intentional falsification or invention of data or citation in an academic exercise will be considered a violation of academic integrity. a. Inventing of altering data for a laboratory experiment or field project.
of grader evaluation error, when, in fact, the work has been altered from its original state. 3. Plagiarism: Plagiarism is the theft and subsequent passing off of another's ideas or words as one's own. If the words or ideas of another are used, acknowledgement of the original source must be made through recognized referencing practice.
footnote citation and by either quotation marks or appropriate indentation and spacing.
merely recast in the student's own words, proper acknowledgement must, nonetheless, be made. A footnote or proper internal citation must follow the paraphrase material. 4. Other Types of Academic Dishonesty:
express permission;
instructor; d. Changing academic records outside of normal procedures;
without the knowledge and consent of the instructor. The above information is taken directly from the S Campus and the Academic Affairs Unit of the Student Senate in conjunction with the Academic Standards Committee. APPENDIX A: ACADEMIC DISHONESTY SANCTION GUIDELINES VIOLATION RECOMMENDED SANCTION (assuming first offense) Copying answers from other students on exam. F for course. One person allowing another to cheat from his/her F for course for both persons. exam or assignment. Possessing or using extra material during exam F for course. (crib sheets, notes, books, etc.) Continuing to write after exam has ended. F or zero on exam. Taking exam from room and later claiming that the F for course and recommendation for instructor lost it. further disciplinary action (possible suspension). Changing answers after exam has been returned. F for course and recommendation for disciplinary action (possible suspension). Fraudulent possession of exam prior administration. F for course and recommendation for suspension. Obtaining a copy of an exam or answer key prior to Suspension or expulsion from the administration. university; F for course. Having someone else take an exam for oneself. Suspension or expulsion from the University for both students; F for course. Plagiarism. F for the course. Submission of purchased term papers or papers F for the course and recommendation done by others. for further disciplinary action. (possible suspension) Submission of the same term papers to more than F for both course. one instructor where no previous approval has been given. Unauthorized collaboration on an assignment. F for the course for both students. Falsification of information in admission application evocation of university admission (including supporting documentation). without opportunity to apply. Documentary falsification (e.g., petitions and suspension or expulsion from the supporting materials medical documentation). university; F for course when related to a specific course. Plagiarism in a graduate thesis or dissertation. Expulsion from the university when discovered prior to graduation; revocation of degree when discovered subsequent to graduation. Please refer to Trojan Integrity: A Faculty Desk Reference, for more information on assessing sanctions. You may also consult with members of the Office of Student Judicial Affairs and Community Standards at any point in the process, (213) 7406666 Note: The Student Conduct Code provides that graduate students who are found responsible for academic integrity violations may be sanctioned more severely than Appendix A suggests. 