A Methods section is a recipe. Cook from one, then reverse one.

Your week07-methods/ folder

Pull your private repo in VS Code (Source Control → Sync). If week07-methods/ is there, you are set. If not, make the folder and save these three files into it:

• rt.html, the Week 2 reaction-time template, unchanged. You turn it into the Stroop task.
• README.md, your report, with the Methods headings.
• analyze_stroop.py, for Part 2.

Right-click → Save Link As. Keep the file names.

Write the Methods, then build what it says

1. Before touching code, fill in Participants, Materials, Procedure, and Data handling in the README. Numbers, not adjectives: which words and ink colours, which keys, how long the fixation and the gap between trials last, how many trials per condition, how the order is shuffled, which trials are thrown out. Commit. The commit time shows the Methods came first.
2. Open week07-methods/rt.html and turn it into the task your Methods describes. The pieces you need:

   // (a) colours, words, and keys at the top
   const colours = [
     { name: "RED", key: "z", rgb: [230, 70, 60] },
     { name: "BLUE", key: "x", rgb: [80, 140, 255] },
     // ...
   ];
   
   // (b) a shuffled list of trials, half congruent, half incongruent
   trials.push({ condition: "congruent", word: ink.name, ink: ink });
   trials.push({ condition: "incongruent", word: other.name, ink: ink });
   trials = shuffle(trials);
   
   // (c) in draw(): show the word in the ink colour instead of the square
   fill(current.ink.rgb[0], current.ink.rgb[1], current.ink.rgb[2]);
   text(current.word, width / 2, height / 2);
   
   // (d) in keyPressed(): send z/x/n/m to handleResponse(k),
   //     and work out correct = (k === current.ink.key) ? 1 : 0
   
   // (e) add condition, word, ink, key, correct to results.push({...}),
   //     to the CSV header, and to the .join(",") list

3. Every time the code makes you decide something the Methods did not say (does the word stay until a key press? what happens after a wrong key? is a practice block counted?), write it under Decisions I had to invent, then add it to the Methods.
4. Run a practice block, then the scored block on yourself. Export stroop-week07.csv into week07-methods/. It needs the columns condition, rt, and correct. Write your congruent mean, incongruent mean, and Stroop effect under My result.

The test: could a stranger rebuild your page from your Methods alone, without seeing the code?

From Hedge's result back to Hedge's Methods

Here is what the paper reports for the Stroop task. Every participant was tested twice. Reliability is an intraclass correlation (ICC): 1 means people kept exactly the same rank order from session 1 to session 2, 0 means no relation.

1. Guess first. Under My guess in the README, write the Methods you think produced this: who, how many trials, which colours and keys, how long between sessions, which trials and which people were excluded. Commit.
2. Reveal. Open the open-access PDF and read the Stroop Methods (or open the box below). Under What I missed, list three details you got wrong or never thought of, and whether each could change the result.
3. Reanalyse. In a terminal in week07-methods/:

   python3 analyze_stroop.py --mine stroop-week07.csv

   It downloads the tidied Hedge data (every trial, both studies, both sessions), applies their exclusion rules, and prints the group effect, the session 1 vs 2 correlation, and where your own effect falls. Paste output/numbers.md into the README.
4. Run it again with --rule course (our 150–2500 ms rule instead of theirs). What changed? One sentence on why exclusion rules belong in a Methods section, written in numbers.