Doing Computational Social Science

John McLevey
Sage, London, 2022

Description

Computational approaches offer exciting opportunities for us to do social science differently. This beginner's guide discusses a range of computational methods and how to use them to study the problems and questions you want to research.

It assumes no knowledge of programming, offering step-by-step guidance for coding in Python and drawing on examples of real data analysis to demonstrate how you can apply each approach in any discipline.

The book also:

• Considers important principles of social scientific computing, including transparency, accountability and reproducibility.
• Understands the realities of completing research projects and offers advice for dealing with issues such as messy or incomplete data and systematic biases.
• Empowers you to learn at your own pace, with online resources including screencast tutorials and datasets that enable you to practice your skills and get up to speed.

For anyone who wants to use computational methods to conduct a social science research project, this book equips you with the skills, good habits and best working practices to do rigorous, high quality work.

Contents

Introduction - Learning to do computational social science

Part I: Foundations

1. Setting up your open source scientific computing environment
2. Python programming: The basics
3. Python programming: Data structures, functions and files
4. Collecting data from Application Programming Interfaces (APIs)
5. Collecting data from the web (Scraping)
6. Processing structured data
7. Visualisation and exploratory data analysis
8. Latent factors and components

Part II: Fundamentals of Text Analysis

9. Processing natural language data
10. Iterative text analysis
11. Exploratory text analysis
12. Text similarity and latent semantic space

Part III: Fundamentals of Network Analysis

13. Social networks and relational thinking
14. Connection and clustering in social networks
15. Influence, inequality and power in social networks
16. Going viral: Modelling the epidemic spread of simple contagions
17. Not so fast: Modelling the diffusion of complex contagions

Part IV: Research Ethics and Machine Learning

18. Research ethics, politics and practices
19. Machine learning: Symbolic and connectionist
20. Supervised learning with regression and cross-validation
21. Supervised learning with tree-based models
22. Neural networks and deep learning
23. Developing neural network models with Keras and Tensorflow

Part V: Bayesian Machine Learning and Probabilistic Programming

24. Statistical machine learning and generative models
25. Probability: A primer
26. Approximate posterior inference with stochastic sampling and MCMC

Part VI: Bayesian Data Analysis and Latent Variable Modelling with Relational and Text Data

27. Bayesian regression models with probabilistic programming
28. Bayesian hierarchical regression modelling
29. Variational Bayes and the craft of generative topic modelling
30. Generative network analysis with Bayesian stochastic blockmodels

Part VII: Embeddings, Transformer Models and Named Entity Recognition

31. Can we model meaning?: Contextual representation and neural word embeddings
32. Named entity recognition, transfer learning and transformer models