API209
Summer Math Camp

Academic Year: 2026

Active Course Version

Summer 2026

This version of math camp is organized as a compact working environment for the semester ahead. It is less a static archive and more a weekly operating system: lesson material, slide decks, hands-on pages, labs, and a new module on AI-assisted analysis.

Core themes

Wrangling, visualization, modeling, and reproducible reporting.

New in 2026

An explicit AI workflow week built around verification, documentation, and fall-semester use cases.

Design goal

Fewer scattered pages, faster navigation, and clearer week-by-week entry points.

Teaching Team

Photo of Professor Levy

Dan Levy

Professor โ€ข Harvard Kennedy School

Photo of Teaching Fellow

Rony Rodriguez-Ramirez

Summer TF โ€ข Harvard University

Schedule office hours

What This Camp Covers

Technical skills

  • data cleaning and reshaping with the tidyverse
  • clear plots for exploration and communication
  • introductory model interpretation
  • Quarto-based reproducible workflow

Workflow skills

  • moving from rough code to clean analysis artifacts
  • documenting assumptions and verification
  • using AI tools without surrendering judgment
  • building habits that transfer directly into fall assignments

Weekly Structure

Week 01

Introduction, Context, and Data Wrangling

Build the shared baseline: R habits, wrangling logic, and the working structure of the camp.

Lesson ยท Slides ยท Hands-on ยท Lab

Week 02

Tidy Data and Visualization

Learn how data shape affects analysis quality and how plotting choices affect interpretation.

Lesson ยท Slides ยท Hands-on ยท Lab

Week 03

Modeling and Reproducible Research

Move from descriptive analysis to more defensible workflows for the fall semester.

Lesson ยท Slides ยท Hands-on ยท Lab

Week 04

AI-Assisted Analysis and Research Workflows

Practice using Codex, Claude Code, and related tools for bounded tasks while preserving reproducibility and verification.

Lesson ยท Slides ยท Hands-on ยท Lab

Why AI Is Included

Many students will use AI tools during the fall whether or not the course addresses them directly. The 2026 design treats that as a workflow question, not a novelty topic:

  1. define the task before prompting;
  2. ask for bounded help;
  3. run and inspect the output yourself;
  4. verify one important result manually;
  5. document what changed because of AI.