Course overview
Start hereAI for Statisticians
START HERE

Daniel's Wednesday, Two Ways

8 lessons · 24 prompts

A day in the life of a Statistician: what changes with these prompts.

Track progress as a member

Daniel is a health policy statistician at a state agency.

Daniel starts Wednesday with a request from his director: find out whether weekend clinic hours changed patient wait times. The ask is vague, so he opens ChatGPT and uses the framing prompt. He turns it into a clear question, a plan, and a note about what sample size would be needed.

Next he opens a messy spreadsheet from the scheduling team. It has missing insurance codes and duplicate visit IDs. With Claude, he uses the cleaning prompt to draft a reproducible script, then checks each step on ten rows before running it on the full file.

After lunch he runs a test comparing wait times at two clinics. He uses the running statistical tests prompt in Gemini to list assumptions and read the output. The result is not dramatic, so he writes a short note about uncertainty himself.

Before his stakeholder meeting, he uses the reports prompt in ChatGPT to draft a one-page summary and a few likely questions. He edits the language until it sounds like him. With the time he saved, he leaves on time and makes dinner with his family.

Before

  • Requests arrive vague and urgent.
  • Data cleaning eats the morning.
  • Tests and charts wait until late.
  • Explaining results feels rushed.

After this course

  • Clear questions start the day.
  • Cleaning scripts run with fewer surprises.
  • Tests and charts get proper review.
  • Stakeholder talks feel calm and honest.

What you'll learn

  • Frame questions: Turn a vague request into a hypothesis, an analysis plan, and a sample size estimate with AI help.
  • Design studies: Draft survey items, spot leading wording, and plan experiments that can answer the question.
  • Clean data: Create reproducible cleaning steps, catch errors, and document every variable change.
  • Run tests: Choose a suitable test, check assumptions, and read the output with a careful eye.
  • Build models: Plan features, write training code, and find out why a model fails.
  • Explain uncertainty: Translate results into effect sizes, stress-test conclusions, and prepare honest insights.
  • Make charts: Pick the right chart, write plotting code, and improve clarity without hiding uncertainty.
  • Report clearly: Draft report sections, write plain-English summaries, and get ready for stakeholder talks.

How this course works

  1. 8 lessonsOne task of your job each, from framing statistical questions to reports and stakeholder talks.
  2. Ready-to-paste promptsCopy, fill in the parts in {{brackets}}, paste into ChatGPT, Claude or Gemini.
  3. Tick and completeTick the prompts you tried and mark each lesson complete.
  4. Get certifiedFinish and keep the prompts as your own library.