Make a project something a co-author could rerun. Practice on a messy one, then do one of your own.
A paper you’re writing with Kari Nordmann: Monthly or yearly? Price framing and willingness to pay. A survey experiment, cleaned in Excel, written up in Word. Fictional, and completely typical.
Start the agent in an empty folder, and paste the download prompt from the handout, section 11. (Or copy practice/messy-project from the course files.)
messy-project/
├── survey_export_2026-03-14.csv the raw survey, 240 rows
├── data_clean_FINAL.csv cleaned by hand in Excel
├── analysis.py first try
├── analysis_v2.py the analysis
├── table1.txt results
├── table1 (copy).txt results?
├── notes.txt "the crazy WTP ones?? ask Kari"
└── draft_v3.docx the Word draft
Could Kari rerun this? Could you, next spring?
| Invisible | Visible | |
|---|---|---|
| Writing | Word: you can’t see what changed | Text: every change shows |
| Data | Clicking in Excel: no record | A script: the steps are the record |
−0.1% → +2.2%
Average growth when public debt is above 90% of GDP: as published, and after a grad student redid it from their spreadsheet in 2013.
AVERAGE over rows 30–44, not 30–49. Five countries left out.Your own rerun finds it first.
Rerun everything from scratch. Do you get the same numbers?
The practice folder was cleaned by hand in Excel, with two rules. Drop the speeders: written down. Drop three answers above 5,000 NOK: not written down.
| N | Effect | p | |
|---|---|---|---|
| Keep the three | 223 | −292 NOK | 0.26 |
| Drop them | 220 | +53 NOK | 0.02 |
An undocumented click in Excel decided the sign of the result. The rerun test found it.
A wrong script you can read beats a right click you can’t remember.