Computing environment · PSYCH 302 / 305

Repository, Codespace, and AI disclosure

If Week 0 is still incomplete, submit the GitHub form and accept the private-repo invite. Week 1 (due Tuesday 8 September, 17:00) requires all four: a repository you create, GitHub Pages on that repo, a fork of the whole class handbook (kylemath/Psych302-305-KyleMathewson) with your page as that fork’s index.html, and a pull request from the fork back to the class repo. Paste those four links on Canvas — a live page is not the submission. Do not fork only your new repo. The instructor will merge the coolest pull request as the new default homepage.

Repository

One folder for the term

Week 1: (1) new repo with index.html, (2) GitHub Pages on that repo, (3) fork the entire class repo and replace only index.html, (4) open a pull request back to the class repo (the coolest one becomes the new class homepage). Later weeks: reports in lab-notes/, data in data/, generated papers in report/.

Codespace

Editor in the browser

On your repository: Code → Codespaces → Create. Use the Simple Browser or port 8765 to open the laboratory page. Prefer a 2-core machine. No software need be installed on your laptop.

AI use

Disclose every week

You may use Copilot or an editor agent to explain a line or draft text. You stay the author. You may not paste another student’s data into a prompt. Each report states whether you used it, what you asked, what you kept, and what you changed. If none is available, write “no Copilot tonight.” Do not purchase a plan.

Alternatives (in order)

Codespace — preferred.

github.dev — press period (.) on the repository page, or replace github.com with github.dev.

ZIP download — edit locally; push later with assistance if needed.

Open the HTML file — use this handbook in a browser; complete the report by Tuesday.

Agentic coding

An agent can propose an edit, run a command, and show you the result. That does not make the report its work. You choose the question, you keep the files, you read the output, and you decide what stays. The same loop appears in Week 4 without any assistant: edit a file, run a script, read the paper, edit again.

Strategies that work in this course

Small asks. “Exclude RTs under 150 ms in summarize.py” is better than “write my lab.”
Name the file. Point at report/REPORT.md or data/week02.csv. Do not paste the table into the chat.
Run what it wrote. If you cannot run it, you do not yet have a result.
Read the number. If n is 0 or the mean is 12 ms, the script is wrong. Fix the file, not the sentence.
Keep a trail. A commit, or two sentences in the report, of what you asked and what you changed.

What we do not do

Paste classmate rows, names, or a whole CSV into a prompt.
Accept a bibliography or a mean you have not opened or recomputed.
Buy Copilot, Cursor, or any paid plan for this course.
Submit a page you cannot explain. If you cannot say what the script excluded, redo that part by hand.

Week 1: we discuss this after the playground. Week 4: report.html is the first graded loop. Later weeks: same disclosure on every report.

Commit, push, and pull

Week 1 is the first test: create a repo, turn on Pages, fork the whole class repository, make your page its default index.html, then open a pull request. The coolest one becomes the new class homepage. Commit history is not graded.