IR602: Quantitative Analysis for International Affairs Frederick S. Pardee School of Global Studies Spring 2018 Course Syllabus

Transcription

1 The Formalities: IR602: Quantitative Analysis for International Affairs Frederick S. Pardee School of Global Studies Spring 2018 Course Syllabus Course Instructor: Mahesh Karra Instructor Office Hours (at 152 Bay State Road, Room G04C): Tuesdays and Thursdays, 10:00 AM 11:00 AM Wednesdays, 11:00 AM 12:00 PM Teaching Assistant: TBD Course Website: Log into Blackboard Learn website at learn.bu.edu Class Times and Location: Tuesdays and Thursdays: 11:00 AM 12:15 PM Room XXX The Course: Course Summary: This course presents the principal basic and multivariate statistical methods that may be used in the field of international policy analysis, both for conducting statistical analysis and for understanding impact evaluation studies that are published in advanced research journals. The course aims to make students more effective users of statistical tools for analyzing public policy issues and when suggesting solutions. We will also discuss the interaction between quantitative reasoning, international policy analysis, and applied decisionmaking. This course will prepare international affairs practitioners with statistical reasoning tools and techniques and will emphasize hands-on learning using real social, political, and policy data. Over the semester, we will apply various statistical tools to evaluate causation in international events and policy. We will conduct statistical exercises that compare data and groups before and after an international policy change, political event, or experimental condition. We start with basic and descriptive statistics, including the logic of data visualization. We then focus on causal inference and hypothesis testing, eventually applying methods of inference to policy problems using causal analysis. Finally, we will also discuss methods of data collection, both qualitative and quantitative, and will discuss how to develop appropriate data collection tools for effective monitoring and evaluation. While the course material is mathematical in nature, IR602 should not be seen primarily as a math class. Rather, it focuses on applications of quantitative analysis techniques to issues and problems that Pardee School MA students will encounter in subsequent coursework and in their careers as 1

2 international relations practitioners. Cases and assignments will address the usefulness and limitations of quantitative analysis of actual policy-relevant datasets. Prerequisites and Corequisites: Graduate standing in the Pardee School or permission of instructor. A math background, particularly in probability, statistics, and the fundamentals of calculus, is helpful and encouraged. In the first week of class, I will hold an extra (optional) lecture that covers some fundamental mathematical topics for students who seek to refresh their skills. Although the topics covered in the class are of a more conceptual than mathematically explicit nature, I encourage students to brush up on algebraic and calculus methods throughout the course so as to better grasp some of the less intuitive subtleties. From time to time, I will show the theoretical proofs of various statistical and econometric notions, and thus I expect the students to be able to follow any mathematical steps that are being utilized. Primary Textbook: The course will be built around lecture notes which will be posted on the course website at the beginning of each week. The main reference books for the course are Wooldridge s Introductory Econometrics: A Modern Approach (Wooldridge, 2003) and Howell s Fundamental Statistics for the Behavioral Sciences (Howell, D.C., 1999). Slightly more advanced textbooks, such as Greene s Econometric Analysis (Greene, 2008), Hayashi s Econometrics (Hayashi, 2000), and Wooldridge s Econometric Analysis of Cross Section and Panel Data, also cover most of the topics discussed in class. For students who need to re-familiarize themselves with some of more basic concepts of statistics and econometrics, I recommend to also look at Learning and Practicing Econometrics by Griffiths, Hill and Judge (1993) and Basic Econometrics by Gujarati (1995). I will assign additional readings and small Stata based problem sets each week which will be posted on the course website and discussed in the weekly review sections. Evaluation: Your grade in the class will be derived from the following components: Evaluation Amount in the Semester Problem Sets, p 8 Midterm Exam, M 1 Final Exam, F 1 Your final grade, G, will be determined as follows: M + P 2F + M F + G = max (F,, 2 ) 3 2 where P is the composite (average) grade of the 8 problem sets, i.e. P = 1 p 8 i=1 i. This derivation implies that you cannot do worse in terms of your grade for the entire term than what you get on your final exam. All assignments are to be handed in on the due date in the beginning of class (4:00 PM). Unless otherwise instructed, assignments are to be individually completed. No late work will be accepted. 8 2

3 Grading Policy: The following grading system will be utilized: 85 to 100 = A 80 to 84 = A- 75 to 79 = B+ 70 to 74 = B 65 to 69 = B- 60 to 64 = C+ 55 to 59 = C 50 to 54 = C- 40 to 50 = D Less than 40 = F In the event of decimals, I shall truncate the decimal to the tenth place, round up if the decimal is greater than or equal to 0.5, and round down if the decimal is less than or equal to 0.4. NO EXCEPTIONS! BU Academic Code of Conduct and Policies: All Boston University students are expected to maintain the highest standards of academic honesty and integrity. It is the responsibility of every student to be aware of the university s Academic Conduct Code s contents and to abide by its provisions. Plagiarism and academic dishonesty of any kind will not be tolerated. For additional information, please refer to the complete Academic Conduct Code and the BU CAS Policies and Procedures using the links below. References: 1. Greene, W. H. (2008). Econometric Analysis. Upper Saddle River, NJ: Pearson/Prentice Hall. 2. Griffiths, W. E., R. C. Hill and G. G. Judge (1993). Learning and Practicing Econometrics. New York, NY: Wiley. 3. Gujarati, D. N. (1995). Basic Econometrics. New York, NY: McGraw-Hill. 4. Hayashi, F. (2000). Econometrics. Princeton, NJ: Princeton University Press. 5. Howell, D. C. (1999). Fundamental Statistics for the Behavioral Sciences. 4th Ed. Pacific Grove, CA: Duxbury Press. 6. Wooldridge, J. M. (2002). Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press. 7. Wooldridge, J. M. (2003). Introductory Econometrics: A Modern Approach. Thomson South- Western. 3

4 Course Outline: The speed at which each section will be completed depends on each class. Week numbers are merely approximations and may vary considerably depending on the time constraint. However, all material is expected to be covered by the end of the course. Chapters, unless otherwise mentioned, refer to the required text. I. INTRODUCTION: Week 1 a. Introduction to Programming and Stata: Week 1 i. Stata introduction and tutorial ii. Introduction to the Demographic and Health Surveys iii. Registration to download DHS data II. DESCRIPTIVE STATISTICS, DATA VISUALIZATION, Week 2 a. Descriptive Statistics, Week 2 i. Key statistics of central tendency ii. Population and sample statistics iii. Laws of Large Numbers and Central Limit Theorems iv. Applications using Stata ASSIGNMENT 1 DUE III. INFERENTIAL STATISTICS: Week 3 a. Hypothesis Testing I i. Type 1 and 2 Error ii. Misclassification (Sensitivity and Specificity) iii. Power b. Hypothesis Testing II i. Univariate analyses (z-scores, t-test) ii. Introduction to ANOVA and correlation analyses iii. Balance tables (Table 1) ASSIGNMENT 2 DUE IV. REGRESSION ANALYSES: Weeks 4-5 a. Fundamentals of Regressions: Week 4 i. Ordinary Least Squares ii. Derivation of OLS iii. Regression as a comparison of group means ASSIGNMENT 3 DUE b. Hypothesis Testing and Inference with Regressions, Week 5 i. Univariate Regression Analyses ii. Multivariate Regression Analyses iii. Bias, Consistency, Endogeneity iv. Heteroskedasticity, standard error correction, and design effects 4

5 ASSIGNMENT 4 DUE MIDTERM EXAM V. CAUSAL INFERENCE IN EXPERIMENTS: Weeks 6-7 a. Causal Inference Part I, Week 6 i. Counterfactual Analysis and the Ideal Experiment ii. Causal Diagrams (DAGs) iii. Designing the Ideal Experiment iv. Biases in and Limitations of RCTs b. Inference in Experiments: Week 7 i. Individual Randomized Controlled Trials (RCTs) ii. Cluster-Randomized Controlled Trials iii. Staggered and step-wedged RCTs iv. Sample size calculations for experiments ASSIGNMENT 5 DUE VI. CAUSAL INFERENCE IN QUASI-EXPERIMENTS: Weeks 8-9 a. Causal Inference Part II, Week 8 i. Deviations from the ideal experiment ii. Quasi-experimental methods for impact evaluation iii. Non-experimental methods for impact evaluation b. Quasi-Experimental Methods: Week 9 i. Regression Discontinuity Design ii. Instrumental Variables iii. Difference-in-Differences, Pooled OLS, Introduction to Fixed Effects ASSIGNMENT 6 DUE VII. INFERENCE II NON-EXPERIMENTAL METHODS: Week 10 i. Matching methods (PSM, Coarsened Exact Matching) ii. Cross-Sectional Approaches iii. Other Regressions: Binary Dependent Variables (Logistic, Probit) ASSIGNMENT 7 DUE 5

6 VIII. DATA COLLECTION AND RESEARCH DESIGN: Weeks i. Primary data collection methods ii. Survey / questionnaire development iii. The ethics of primary data collection and human subjects research iv. Introduction to CommCare ODK v. Introduction to qualitative research 1. The qualitative approach, purpose, and objectives 2. Designing and implementing qualitative studies (IDIs, FGDs) 3. Analysis of qualitative data 4. Integrating qualitative and quantitative methods (mixed methods) ASSIGNMENT 8 DUE FINAL EXAM 6