Linear Models (405-0-1)

Instructors

Saera Lee

Meeting Info

Scott Hall 212: Tues 9:00AM - 11:50AM

Overview of class

This course is about linear models, the major workhorses of statistics for description and prediction, and one of the most common quantitative methods in political science. We will use a linear models framework to discuss significance tests, graphical displays, tests of assumptions, interpretation of coefficients and interactions, and questions of causal inference. We will also work through statistical computing skills such that students can use all of the above in their own work.

Registration Requirements

Recommended pre-requisite classes:
Political Science 403 or equivalent

Learning Objectives

This course introduces students to linear regression models for the analysis of quantitative data and provides a basis of knowledge for more advanced statistical methods. After covering the classic normal linear regression model and its assumptions, we will explore the consequences and remedies for violations of these assumptions, including omitted variables, heteroscedasticity, autocorrelation, and endogeneity. We will also explore the use and interpretation of continuous, ordinal, nominal, and indicator variables as well as interactions between them. If time permits, we will also discuss missing data and basic models for limited dependent variables. Along the way and primarily in the lab session, students will learn the basics of data collection, organization and management; measurement; data visualization and display; and univariate, bivariate and multivariate descriptive statistics.

Teaching Method

Lecture

Evaluation Method

Homework: 40%
Exams: 40%
Replication paper:20%

Enrollment Requirements

Enrollment Requirements: Reserved for Graduate Students.

Associated Classes

DIS - Locy Hall 213: Fri 11:00AM - 11:50AM

DIS - NO DATA: NO DATA