Introduction to Statistical Modelling

We live in an age of data abundance. We know more about the social and political world than any time before in history. Making sense of all these data is an increasingly valuable and marketable skill for social scientists. This class seeks to lay solid foundations for working with quantitative data, whether it is public opinion polls, regional estimates of economic growth, or casualties in civil wars (etc). Relying on best practices from data science, it will devote considerable attention to simple yet crucial questions like “How can you import and clean messy, real world data before analysis?”; “How should you set up a virtual environment to ensure that other people (e.g. future you) can make sense of your analyses?”, “How can you communicate quantitative insights to a lay audience?”. While the course teaches R – an increasingly popular and free software for data analysis – most of the practical lessons from the class are directly transferable to alternative data analytical environments (from Excel through Tableau to Python). The course will be organized in three parts: 1) Data carpentry: Setting up the analytical environment, importing and cleaning data before analysis. 2) Data visualization: Creating figures is the most effective tool to explore and understand your data and also to communicate insights to a lay audience. 3) Modeling: Create effective summaries of the data, explore relationships and quantify uncertainty. The objective of the course is to prepare students to be able to perform fundamental statistical analyses in R on their own as part of their MA thesis or at a. future job.