Resources
The required text, its companion website, computing setup, and course templates.
The required text
Ma, J. (forthcoming). Computational & AI-Assisted Methods for Social Sciences: A Research Design Primer. SAGE.
Assigned chapters are posted in Canvas. You do not need to buy anything. The weekly schedule names the chapters for each week, and each weekly guide says what to look for in them. Contact the instructor if you cannot open a chapter.
The book explains research tasks independently of any programming language. Its 12 chapters come in two kinds, and the course uses them differently.
| Chapters | What they do | How we use them |
|---|---|---|
| Theory: 1, 2, 3, 5, 7, 9, 11, 12 | Build the reasoning: what a measure commits you to, what a method can and cannot claim, what makes a result trustworthy | Reading, the study presentations, and the reflections |
| Exercise: 4, 6, 8, 10 | Put the same reasoning to work on your own data, one deliverable at a time | The exercises that make up Milestones 2 through 5 |
The companion website
The book’s free companion website holds everything that changes faster than a printed book: setup instructions, code walkthroughs, tool lists, templates, and data. We link to it throughout the course instead of copying it. Start from the page for the chapter you are working on.
| Stage | Theory chapter and its studies | Exercise chapter and its walkthrough | Primers and tools |
|---|---|---|---|
| Project setup | Chapters 1–2: project templates and a recorded example | Getting started | |
| Data management | Chapter 3: studies annotated for data management | Chapter 4: retrieving a corpus from OpenAlex | Working with text data · data tools |
| Concept representation | Chapter 5: studies annotated for concept representation | Chapter 6: dictionary, prompt, classifier, and audit | Prompting and its failure modes · Supervised learning, gently · coding tools |
| Data analysis | Chapter 7: studies annotated for computational analysis | Chapter 8: networks, simulation, prediction, embeddings | Supervised learning, gently · analysis tools |
| Scientific communication | Chapter 9: studies annotated for scientific communication | Chapter 10: frozen environments, reports, dashboards, uncertainty | communication tools |
Chapters 1, 11, and 12 have no page of their own and use the shared materials. Three collections cut across the chapters:
- CSS Empirical Studies Database: published computational studies, annotated by design stage. The four theory-chapter pages filter it by stage. It is your reference for the study presentation, which covers two recent studies that are not in it.
- Worked notebooks and data: text representation, topic modeling, and network analysis with saved outputs, plus a sample co-citation network for practicing Chapter 8’s network steps.
- Chapter materials and the library: the same holdings indexed by chapter and by task.
If you find an error in the book or on the companion website, the errata page says how to report it.
Computing setup
Follow the companion’s getting-started guide. It offers two routes: a browser notebook that needs no installation, and a local environment for data you cannot upload or a setup you want to keep. It also explains how to get access to a language model through your institution, a commercial provider, or a model that runs on your own computer.
The companion’s examples are in Python. R is welcome if you can complete and document the same steps; the guide and the tool pages include R options. Whatever you choose, record it: the language and package versions, and the model name, version, and settings, are part of your method. Milestone 1 asks for your route, language, and model access. Add versions and settings as soon as you run something.
Before Wednesday of Week 1, try the guide’s first run and keep a note of anything that fails. Setup and troubleshooting count toward the weekly budget, so bring a saved error to class before it costs you the week.
Several exercises call a language model. You do not need a paid AI subscription for this course. Check what Indiana University provides and what its terms allow for your data, or use a local model, and ask the instructor if neither works for you. Do not upload restricted or confidential material to a hosted notebook or AI service.
The project skeleton
Download the Chapter 2 project skeleton and use it as your project folder. The Chapter 2 page shows a rules file, a current-state file, and a promptlog entry filled in for a worked example, and includes a recording of the structure steering a real analysis. The skeleton supplies the templates you will use most:
-
AGENTS.md: the rules an agent follows in your project, including what it must verify. -
memory/CURRENT.md,memory/DECISIONS.md, andmemory/PROMPTLOG.md: where the project stands, what you decided, and how your prompts changed. -
kb/README.mdandkb/codebooks/: a map of your materials, and the codebooks and promptbooks that define your measures.
These files are also your record of AI use, so there is no separate log to keep. Using AI explains what goes where.
Course templates
Two short documents are specific to P539:
- Reflection guide: how to write a reflection, and which project files to draw on for each one.
- Milestone 5 checklist: what Milestone 5 should contain before you submit it.
Everything else you need is in the book, in the skeleton, or on the assignments page.
More help
The companion’s resource list points to the readings the book builds on, and to courses and data sources beyond it. IU Libraries Research Data Services can help with data management. For course-specific help or an office-hours appointment, email Ji Ma.