Materials
Course Overview and learning objectives
This course aims to equip economists who regularly use Stata and other modern statistical programming languages to leverage AI-assisted coding tools effectively and responsibly. It introduces participants to AI coding assistants (GitHub Copilot) within the Positron IDE, across two 3-hour sessions with a self-study period in between and afterwards.
By the end of this course, participants will be able to:
Understand
- Core LLM concepts (tokens, prompts, context windows, agents) and how they affect AI-assisted coding
- Strengths and limitations of AI code assistants for analytical work
Apply
- AI assistant for 3R: read, revise, and reuse existing code.
- Prompts engineering and context management for precise AI answers and code generation
- Planning and executing complex tasks with multi-step agentic workflows
#tools,skills, and/agents- Supervise and course-correct AI-generated code through iterative review
Take control
- Apply data-security practices when using AI tools with sensitive or confidential data
- Follow responsible-use guidelines for AI-assisted research workflows
Prerequisites
Complete these setup steps before the first session:
Day 1 — Introduction to AI-Assisted Coding
3-hour in-person / hybrid / online session
| Time | Topic |
|---|---|
| 10 min | Welcome, introductions, course and materials overview |
| 30 min | Software overview: Positron, GitHub Copilot, and Stata setup |
| 40 min | AI in action with Stata (R): my typical data workflow using AI |
| 10 min | Break |
| 30 min | AI overview: how GitHub Copilot and LLMs work, key concepts, and capabilities |
| 30 min | Cookbook: securing sensitive data, guardrails, and responsible AI use |
| 30 min | Self-study exercises overview and Q&A |
Self-Study
2–3 days of independent practice between sessions
Independent practice is meant to help you build confidence and discover how to use AI tools in your own workflow. You can pick any case study that interests you. You are encouraged to try the AI and report your observations, thoughts and questions using Team’s chat.
- If possible, your challenges will be answered immediately by the team or other participants.
Day 2 — Everyday AI-Assisted Coding
3-hour in-person / hybrid session
| Time | Topic |
|---|---|
| 20 min | Q&A from self-study: challenges and discoveries |
| 30 min | Exercise: planning and executing complex tasks with parallel agentic workflows |
| 30 min | Context engineering: #tools, skills, MCP, and other relevant features |
| 10 min | Break |
| 45 min | Exercise: using the right #tools, integrating AI skills, developing /prompts and /agents |
| 30 min | Q&A and discussion |
| 5 min | Feedback survey |
| 10 min | Closing remarks and next steps |