Y577: Political Data Analysis II (Contextual Analysis)
This course builds upon the linear regression model with various empirical approaches that are useful for the study of contextual political factors. The emphasis will be on developing practical proficiency with the estimation and interpretation of statistical models, but students will also learn how "context" has been conceptualized across different subfields of political science. We will start with a few selected multilevel analysis methods to analyze contextual factors embedded in grouped data. We will then study spatial regression models that incorporate geographical proximity as a way to analyze the contextual importance of physical space. The second half of the semester will focus on the study of networks to understand recent network-analytic research in political science.
Year(s) Offered: 2017
Semester(s) Offered: Spring
Class time: Wednesday : 1-3pm
Capacity: 10 (spring 2016)
Instructor: Armando Razo - firstname.lastname@example.org
Other Contact(s): Armando Razo - email@example.com
Prerequisites: Y575 or equivalent (e.g., S501) is required.
Algebra Required?: Yes (for notation); No (for proofs) No (for homework)
Calculus Required?: Yes--occasionally (for concepts); No (for derivations); No (for homework)
Day(s) per week offered: two-hour weekly meeting will include mini-lectures and statistical exercises in a computer lab setting
Recommended follow-up classes: S503; S651 (Social Network Analysis); S681 (Spatial Statistics)
Substantive Orientation: social sciences
Statistical Orientation: observational data analysis, statistical computing
Books used: There are not required books. Our main textbook will be an unpublished book manuscript. Articles and other readings will be available online.
Applied/Theoretical: This is a mostly an applied course. Y577 gives a conceptual overview of relevant statistical models, but the focus is on estimation and interpretation.
Formal Computing Lab?: Yes
Software Used: R, Stata
How the software is used: Software will be mostly used for data analysis with some minimal programming tasks (scripts for data analysis).
Problem Sets: Yes
Data Analysis: Yes
Presentations: Yes, students will be required to either formulate a research design for their own projects or otherwise attempt to replicate a published article
Exams: Two take-home exams
Keywords: contextual, hierarchical, multilevel, HLM, grouped, spatial, multilevel, networks, SNA, interdependence
Comments: Syllabus will be available in Canvas site the week before classes start.