• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site
Master 2019/2020

Methodology and Research Methods of Political Science

Type: Compulsory course (Applied Politics)
Area of studies: Political Science
When: 1 year, 2, 3 module
Mode of studies: Full time
Instructors: Evgeny Sedashov
Master’s programme: Applied Politics
Language: English
ECTS credits: 4

Course Syllabus

Abstract

This course serves as an introduction to quantitative political methodology. We will first cover the general issues related to research design in political science. We will discuss problems of measurement and operationalization, validity and reliability of measurements, and the basics of writing research papers. After that we will proceed to the discussion of statistical methods and linear regression. We will start from first principles and gradually build skills required for thorough understanding of quantitative methods used in modern political science. Finally, we will also cover some of the topics from causal inference. The class strives to maintain the balance between building theoretical understanding and practical implementation of specific methods. I strongly believe that one is impossible without the other. On the one hand, understanding theoretical underpinnings of a specific method is pivotal for realizing the limits of its use. On the other hand, the theory separated from practical implementation tends to be a bit dry, so getting acquainted with data management and actual implementation of different models is also important.
Learning Objectives

Learning Objectives

  • Know the basics of probability theory
  • Know fundamental concepts and ideas from mathematical statistics
  • Know major components of political science research
  • Being able to build research designs and write research papers that address questions from political science and economics
  • Know the theory behind OLS regression and its main assumptions
  • Being able to perform basic data management tasks such as merging and reshaping datasets
  • Being able to implement OLS regression with the tools of Python or R language
  • Being able to perform basic visualizations that illustrate results of regression analysis and summary statistics
  • Know main maximum likelihood models that are frequently used in political science and economics
Expected Learning Outcomes

Expected Learning Outcomes

  • Know the principles of scientific research
  • Know the main steps of scientific inference
  • Know basic logical components of theory building
  • Know the main steps of research design in political science
  • Know how to find research designs that best fit research questions and theory
  • Know the stability assumption (SUTVA)
  • Able to analyze experimental designs for their correspondence to SUTVA
  • Know criteria for classification of assignment mechanisms
  • Know the definition of assignment mechanism
  • Able to classify assignment mechanisms in accordance with individualistic, probabilistic and unconfounded properties
  • Know the definition of probability space and substantive interpretation of the term
  • Know Kolmogorov axioms and definition of event
  • Know the concepts of conditional probability and joint probability
  • Know the law of total probability
  • Know the basics of data managements in python, such as reshaping and merging datasets
  • Know the definition of a random variable
  • Know the difference between discrete and continuous random variables
  • Able to derive moments of random variables from moment generating functions and characteristic functions
  • Know the concepts of covariance and correlation
  • Know how to derive distributions of sums of random variables
  • Know basic operations with random vectors and random matrices
  • Know the properties of a random sample
  • Know the properties of a sampling distribution
  • Able to construct a random sample in accordance with basic algorithms
  • Know basic concepts of random variable convergence: almost sure, in probability, and in distribution
  • Know the statement of the theorem for the Strong and Weak Laws of Large Numbers
  • Know the statement of the Central Limit Theorem
  • Able to illustrate Laws of Large Numbers and Central Limit Theorem using tools from Python or R
  • Know the basic ideas behind statistical estimation
  • Know the fundamental properties used to evaluate statistical estimators: consistency, unbiasedness, efficiency
  • Know the ideas behind basic approaches to derivation of estimators: Method of Moments, MLE, Least Squares, and Bayes
  • Know the ideas behind difference-in-means test
  • Know how to derive confidence intervals and compute p-values
  • Know how to implement difference-in-means test in Python or R
  • Know how to substantively interpret and illustrate results from bivariate regression
  • Know how to derive OLS estimates for bivariate regression
  • Able to perform bivariate regression analysis with the tools of Python or R language
  • Know the basic ideas behind multiple regression analysis
  • Know how to derive OLS estimates for multiple regression using matrix notation
  • Know Gauss-Markov assumptions
  • Able to perform multiple regression analysis and appropriate visualizations in Python or R
  • Know the full proof of Gauss-Markov Theorem
  • Know the asymptotic properties of OLS estimators
  • Know how to perform hypothesis testing and how to derive confidence intervals in multiple regression settings
  • Know how to deal with violations in standard Gauss-Markov assumptions
  • Know the consequences of violations in standard Gauss-Markov assumptions
  • Able to perform basic tests for violations in Gauss-Markov assumption in Python or R
  • Know the basic families of MLE models
  • Know how to derive logit and probit estimates using MLE framework
  • Able to implement event count and discrete choice models in Python or R
  • Able to substantively interpret results from basic MLE models
Course Contents

Course Contents

  • Introduction
    Introduction. Course overview. Principles of scientific research.
  • Research Design
    Research design in political science. Research questions, principles of theorizing and data collection.
  • Elements of Causal Inference & Rubin Causal Model
    Causal inference and Rubin Causal Model. Statement of the problem that causal inference seeks to solve. Assumptions of causal inference. Potential outcomes. Assignment mechanisms. Stability assumption (SUTVA). Introduction to randomization.
  • Basics of Probability Theory
    Probability spaces. Events. Kolmogorov axioms. Conditional probability. The law of total probability. Independence.
  • Random Variables and Their Types
    Definition of a random variable. Discrete and continuous random variables. Distributions of random variables. Transformations and expectations of random variables. Moments and central moments of random variables. Moment generating functions and characteristic functions.
  • Multiple Random Variables
    Multiple random variables. Joint and marginal distributions. Conditional distributions. Covariance and correlation. The law of iterated expectation. Laws of total variance and total covariance.
  • Populations and Samples
    Random sampling from a population. Properties of a random sample. Sampling distribution and its properties.
  • Central Limit Theorem and the Law of Large Numbers
    Convergence of random variables. Almost sure convergence. Convergence in probability. Convergence in distribution. Strong and weak laws of large numbers. Central Limit Theorem.
  • Statistical Estimators and Their Properties
    Definition of a statistical estimator. Properties of estimators. Unbiased estimators. Consistent estimators. Efficient estimators. General approaches to derivation of estimators.
  • Difference-in-Means and Hypotheses Testing
    Difference-in-means test. Confidence interval. Hypothesis testing and p-values. Type I and type II errors.
  • Bivariate Linear Regression and Its Interpretation
    OLS estimation of bivariate regression. Interpretation of bivariate regression results. Implementation of bivariate regression in Python.
  • Multiple Linear Regression and Gauss-Markov Theorem I
    Introduction to multiple regression. Multiple regression in matrix form. OLS estimation of multiple regression. Gauss-Markov Theorem. Implementation of bivariate regression in Python.
  • Multiple Linear Regression and Gauss-Markov Theorem II
    Inference in multiple regressions. Hypotheses testing and confidence intervals. Asymptotic properties of OLS estimators.
  • Multiple Linear Regression – Violations in Assumptions
    Violations of standard Gauss-Markov assumptions. Heteroskedasticity. Model misspecification. Non-zero expectation of error terms.
  • Introduction to MLE models
    Maximum Likelihood Estimation (MLE) in economics and political science. Common families of MLE models. Discrete choice and event count models. MLE estimation of binary choice models.
  • Wrap-up and Final Review
    Wrap-up of the class. Final Exam Review.
Assessment Elements

Assessment Elements

  • non-blocking Домашнее Задание 1
  • non-blocking Домашнее Задание 2
  • non-blocking Домашнее Задание 3
  • non-blocking Домашнее Задание 4
  • non-blocking Домашнее Задание 5
  • non-blocking Домашнее Задание 6
  • non-blocking Домашнее Задание 7
  • non-blocking Домашнее Задание 8
  • non-blocking Промежуточный Экзамен
  • non-blocking Итоговый Экзамен
    Exam takes place on 13 of June, 13.00 – 15.00. Please note that this time is exact: you need to start at 13.00 and end your attempt before 15.00. Only one attempt is allowed. Please, make sure your Internet connection is stable for the duration of the exam. This examination will be conducted in writing, online, with asynchronous proctoring. The platform for the exam is https://hse.student.examus.net Your computer, camera, and internet connection should satisfy the following technical specifications https://www.hse.ru/en/studyspravka/distance_stud_proctor Please login to the system 15 minutes before the examination starting time. The exam will start and end electronically, on time. Examus allows users to test their system before the exam. To take part in the exam, you will have to ◦ login in the system ◦ turn on/check that your camera is working ◦ confirm your identity ◦ test the system ◦ wait for the first question to appear on your screen. During examination, it is prohibited to ◦ leave the room ◦ communicate with other people in the room (if present), or through social media ◦ open any websites other than through links provided in the exam itself (i.e., the link to the calculator) ◦ use a phone or any other smart device ◦ look outside the camera field of coverage for more than 15 seconds During examination, you are allowed to ◦ use the paper to take notes and perform calculations Your desk should be clear of any objects except a mouse/keyboard, and, if you choose so, couple of clean pieces of paper. If your electronic system/connection is interrupted for less than 5 minutes, the examination can be resumed without penalties. If the interruption is longer than 5 minutes, you will have to re-take the exam at a later time and day. The exam will cover the material from the Module 3. Module 2 part of the course was covered in the Midterm and will NOT be covered in this exam. Exam will consist of a number of multiple choice, insert the number, short answer, and True/False questions. Multiple choice questions have only one correct answer. The total number of questions is 25. Each correctly answered question will give you 4% of the exam final grade. Incorrect answers get 0 % score. You will have a maximum of 2 hours (13.00 – 15.00) to complete the exam. Once you started, there will be only one attempt to complete the exam. The questions will appear one by one, with one question for page. Once you answered the question and confirmed the answer, the new question would pop up. You can go back to the question later if you wish to do so BEFORE you submit your attempt. You will NOT know the total score immediately as short answer questions require the manual grading. You will know the partial score. The maximum partial score is 64 %, or 6.4 out of 10. The weight of the final exam grade in the final grade for the entire course is 20%. Make-up final exam will be similar in form and complexity as the original final exam. In addition to the final exam, you have the access to the short practice test at the examus https://hse.student.examus.net platform. You should take this practice test couple of days prior to the actual exam in order to get acquainted with the system. Practice test is always available to you, and you can take it as many times as you want.
  • non-blocking Активность на Занятиях
Interim Assessment

Interim Assessment

  • Interim assessment (3 module)
    0.1 * Активность на Занятиях + 0.06 * Домашнее Задание 1 + 0.06 * Домашнее Задание 2 + 0.06 * Домашнее Задание 3 + 0.06 * Домашнее Задание 4 + 0.06 * Домашнее Задание 5 + 0.06 * Домашнее Задание 6 + 0.06 * Домашнее Задание 7 + 0.06 * Домашнее Задание 8 + 0.22 * Итоговый Экзамен + 0.2 * Промежуточный Экзамен
Bibliography

Bibliography

Recommended Core Bibliography

  • Econometric analysis of cross section and panel data, Wooldridge J. M., 2002

Recommended Additional Bibliography

  • Econometric analysis, Greene W. H., 2000