Overview

As statistical computing and R programming are increasingly supported by AI-assisted tools, this course permits limited use of generative AI (such as ChatGPT, Claude, GitHub Copilot, or similar tools) with the following guidelines:

Permitted Uses (with mandatory disclosure)

You may use AI tools to assist with R programming and statistical computing tasks, including:

To be clear: one goal of this course is to teach you how to use R tools to carry out statistical analyses of categorical data, with well-designed output and graphics for publishable reports. You can’t do this if you simply ask an AI agent to answer a question for you. And, it is a totally different experience for you to think, “How can I write the code to solve this problem?” vs. “How can I ask Claude to do it?”

It is also important that you:

Disclosure Requirement

When you use AI for coding assistance, you must include a brief statement in your submission indicating:

  1. Which AI tool(s) you used (e.g., ChatGPT-4, GitHub Copilot)
  2. What specific problems or tasks you used it for
  3. How you verified the AI-generated code was correct

Example Disclosure Format

AI Disclosure: I used ChatGPT to help debug an error in my ggplot2 code for Figure 2 and to understand how to reshape data using tidyr. I tested all generated code with my own data and verified the output.

Prohibited Uses

You may NOT use AI tools to:

Rationale

The primary learning objectives of this course include developing your ability to interpret statistical analyses, communicate findings clearly, and think critically about categorical data. These skills can only be developed through your own writing and reasoning. While AI can assist with the mechanics of coding, it cannot replace the development of your statistical reasoning and scientific communication skills.

Academic Integrity

Additional Guidance

Verifying AI-Generated Code

When using AI assistance for coding:

  1. Always test AI-generated code with your own data
  2. Verify that the statistical methods are appropriate for your analysis
  3. Understand what each line of code does - do not include code you cannot explain
  4. Check that the output makes sense given your data and research question

Why These Restrictions?

Research has shown that AI tools can sometimes produce incorrect statistical analyses or interpretations. As developing statisticians, you need to build the critical thinking skills to evaluate whether code and analyses are correct. This is only possible if you engage deeply with the material yourself.


This policy may be updated during the term. Any changes will be announced in class and posted to the course website.

## Last updated: December 24, 2025