Fundamentals of Social Data Science in Python

About

A first course on social data science as a science of operationalising social life as data. This is an applied course in data science that moves through some fundamentals of data analysis filtered through theories from social sciences and information sciences.

We will focus on how to make claims, how to consider information lost or gained through data, what is a data structure as a matrix, a graph, and in an embedding space. Focus will be on practical skills and episemic considerations. We will be managing, shaping, and cleaning some data as well as looking at some considerations for claimmaking that come from elementary statistics and vector semantics.

The course will be primarily lab work. Much of this will be group work based.

Key Themes

• The applied use of reason
• Data science as a technique for abstraction and inference Python as a means to manage large data sets
• The web as unstructured data that must be transformed into structured data
• Multiple languages and libraries required for data science
• Considerable exploratory work can be done towards making claims and comparisons
• While confirmatory work is outside the scope of this course, the reasoning behind it is not

Course Objectives

• Writing intermediate level Python including the use of the file system, various data science libraries (mainly pandas), and Pythonic idioms;
• Understand how data can be structured in hierarchical or tabular formats across a variety of file types;
• Be able to wrangle and reshape data in a variety of ways that permit lucid and reasonable comparisons;
• Be able to understand how computational reasoning has been applied to a subset of disciplines focusing on networks and language
• Be able to merge data from multiple sources in order to make original research claims;
• Be able to present data in legible, attractive, and clear graphics;
• Be able to translate computing concepts to mathematical notation.
• Constructing a clear argument for why one particular approach to analysis is more judicious than another paying particular attention to the difference between prediction and explanation (as well as forecasting)
• Writing reports that demonstrate how to shape and merge data to present clear trends or bivariate relationships.

Prerequisites 

This course assumes at least a cursory familiarity with Python and an assumption that the student has at least done rudimentary programming.

Students should use their own laptop for this course. As we are working on introductory skills, virtually any Windows laptop that can run Windows 11 or Mac that can run MacOS Sonoma will be sufficient. Linux compatibility will vary by distribution, but if you are comfortable installing and maintaining Linux, it is likely you will be able to manage the installation of the requisite Python tools. The course instructor uses MacOS and Ubuntu and can informally advise on these systems. Mobile operating systems (iOS, Chromebook, and Android) devices are not sufficient as the specific modules used in the course have not been made easily available on these operating systems.

We further recommend the following:
A harddrive with at least 100GB of space. This is for the downloading and management of video lectures and data sets.
A second screen/monitor as a part of your home workspace. This is because it is useful to have a way to view tutorials or Zoom calls while working on your own material.