Waseda Uni

Waseda University is a large, private university with a main campus located in Shinjuku, Tokyo, Japan. First established in 1882 as the Tōkyō Senmon Gakkō or Tōkyō College by Ōkuma Shigenobu, the school obtained university accreditation and was formally renamed as Waseda University in 1902. The university consists of 13 undergraduate schools and 23 graduate schools. Waseda is one of a select group of top 13 universities assigned additional funding under the Japanese Ministry of Education, Culture, Sports, Science and Technology’s “Top Global Universities” Project.

Waseda consistently ranks among the most academically selective and well-regarded universities in Japanese university rankings.

Course 1: Maximum Likelihood Estimation (22 hours)

Instructor: Dr Daina Chiba

Schedule: 6 – 20 September, 2017

Time: 1:00 PM – 2:30 PM (Mon, Wed, Fri), 1:00 PM – 4:15 PM (Tue, Thu), 10:40 AM – 12:00 PM (20 Sep)

Location: Waseda campus, Waseda University, Japan

Tuition Fees: £300

To apply: Please complete the following online Application form choosing ‘0A’ from the course selection drop down menu

Course 2: Multilevel Analysis (22 hours)

Instructor: Dr Lucas Leeman

Schedule: 13 – 20 September, 2017

Time: 10:40 AM – 4:15 PM (13, 15 Sep), 10:40 AM – 2:30 PM (14, 16, 18, 19 Sep), 10:40 AM – 12:00 PM (20 Sep)

Location: Waseda campus, Waseda University, Japan

Tuition Fees: £300

To apply: Please complete the following online Application form choosing ‘0B’ from the course selection drop down menu

Please Note:
• Students need to make lodging and travel arrangements on their own;
• Waseda cannot sponsor student visa.


Maximum Likelihood Estimation (6 – 20 September, 2017; 22 hours)

Daina Chiba is a Senior Lecturer in the Department of Government at the University of Essex. A graduate of Rice University, he completed his postdoctoral fellowship at Duke University. His research interests encompass the areas of militarized conflict, international institutions, and political methodology. His work has appeared in Political Analysis, American Journal of Political Science, Journal of Politics, Political Science Research and Methods, Journal of Conflict Resolution, and Journal of Peace Research.

Course Content
In this course, students will learn how to build a statistical model to explain the variation of a categorical (binary, ordinal, nominal) dependent variable. They will learn how to build statistical models by properly specifying a likelihood function appropriate to their theory and data. They will then learn how to estimate the unknown parameters of these models using maximum likelihood estimation and how to produce measures of uncertainty (standard errors). Next, they will learn how to use the estimates of the parameters of the model to interpret its substantive implications mainly by calculating substantive effects of the form “my estimates suggest an additional year of education would increase an individual’s chance of turning out to vote by 3%.” Finally, students will learn how to use simulation techniques to put confidence intervals around these substantive effects of the form “my estimates suggest an additional year of education would increase an individual’s chance of turning out to vote by 3%, plus or minus 1%.” Throughout the course there will be an emphasis on how to best describe and explain the models they build and how best to communicate substantive implications to a broad academic audience.

The foundation of building a statistical model is proper development of a likelihood function and that requires an understanding of probability distributions. Thus, we will start with a brief introduction to probability theory at a level appropriate for students with no background in probability theory. The specific models we will subsequently cover are the Bernoulli-logistic model (logit), the normal-linear model (regression), ordered logit, multinomial logit, and event count models (e.g., Poisson, negative binomial).

After finishing this course students should be able to use a wide variety of statistical models in their own work, understand the underlying assumptions of these models, be able to explain the ways in which the models are appropriate or not for the theory and data at hand, and to develop and interpret the substantive implications of the statistical estimates produced by these models.

The course should be taken subsequent to a course on linear regression using OLS. Knowledge of basic calculus will be useful — though not strictly essential. No matrix algebra will be required. That said, statistical models are mathematical models and so we will use a lot of basic algebra and mathematical notation in order to formalize our theoretical intuitions into mathematical (statistical) models. Students should be ready to consume and produce models presented in this way.

Representative Background Reading
Gary King. 1998. Unifying Political Methodology. University of Michigan Press

Statistical Software

Multilevel Analysis (13 – 20 September, 2017; 22 hours)

Lucas Leemann is Reader in political science at the University of Essex. He obtained his PhD from Columbia University where he majored in comparative politics and minored in quantitative methodology. His research in comparative politics focuses on institutional origins and direct democratic institutions. In Data Science he is interested in both measurement (IRT, MrP) and modeling (mostly hierarchical). His articles have been published or are forthcoming in the American Political Science Review, the American Journal of Political Science, the Journal of Politics, Political Analysis, Electoral Studies, and the Swiss Political Science Review.

Course content
This course provides and introduction to multilevel or hierarchical modeling for social scientists. The first couple of days provide the foundational introduction and thereafter we will treat several more advanced topics. The course is suitable for students who have a firm understanding of linear regression and preferably also of binary models. The course will rely heavily on R and Rstudio and also provide a short excursion into Stan. We will also see how the Bayesian approach is helpful for many hierarchical models and especially to generate correct uncertainty measures for all post-estimation quantities.

This course emphasizes the practical use of multilevel models of various types. You will learn how to specify, run and interpret models for a range of different structures, for different data types (continuous and categorical outcome variables). You will learn how to generate uncertainty measures of all possible model-based predictions and to specify Bayesian hierarchical models.

Students will need to be familiar with linear regression and binary models.

Remedial Reading
The best preparation for the course is to ensure that you have the basic skills expected. In case you are not familiar with R/Rstudio please take the free online class at Datacamp before the course begins (https://www.datacamp.com/courses/free-introduction-to-r). We will start with a quick introduction to R, but beginners should start with the Datacamp class. It is also good to take a look at a classic paper on the topic by Steenbergen and Jones (2002, “Modeling Multilevel Data Structures” American Journal of Political Science 46(1): 218-237).

Representative Background Reading
Andrew Gelman and Jennifer Hill. 2007. Data Analysis Using Regression and Multilevel/Hierarchical Models. New York: Cambridge University Press.

Statistical Software
R, Stan