Project Proposal — [Working Title]
STAT 540 · Computing in Statistics · Fall 2026
Two pages. No code required in this file — it is plain markdown and renders with no R, Python or Jupyter.
The proposal exists to catch the one failure this project cannot recover from: an analysis too simple to translate wrongly. One group mean and a scatterplot will translate perfectly, you will find nothing, and there will be nothing to write about. Section 3 is where I check for that.
Submit this and post your dataset and question to [the question board] at the same time.
1 The question
Primary question (one sentence):
…
Why I want to know: one or two sentences.
What I expect to find, before looking: one sentence. Write it down now so the findings section of your report can be honest later.
2 The dataset
| Dataset | e.g. SleepStudy (Lock5Data) / DASL “Hot Dogs” / ACS 5-year Table B19013 |
| Source and vintage | package version, download URL and date, or product + year |
| Collected by, from whom, when | … |
| One row is | … |
| Size | … rows × … columns |
Variables I plan to use — from the codebook, not from guessing:
| Variable | Type | Units / levels | Missing code | What it measures |
|---|---|---|---|---|
What this dataset cannot tell me: one or two sentences.
3 What will make my analysis worth translating
A trivial analysis translates trivially and gives you nothing to find. Your analysis must include at least four of the following, and you commit to which four here, in this proposal — not later, once you have seen what is easy.
Every one of these is a place where two languages can legitimately disagree, and most are also places where both can be wrong together. That is the point.
Tick the ones you will have, and say in a few words where each one appears. “I will probably have some missing values” is not an answer; “Distance is missing for two commuters and it is the denominator of my derived variable” is.
| ✓ | Element | Where it appears in my analysis |
|---|---|---|
| ☐ | Missing values that have to be handled deliberately | … |
| ☐ | A derived variable (a rate, a difference, a recode) | … |
| ☐ | A grouped summary with counts as well as means | … |
| ☐ | A categorical variable with three or more levels | … |
| ☐ | A statistic with a divisor or definition convention — a standard deviation, a variance, a quantile, a correlation | … |
| ☐ | A filter or join that changes the number of rows | … |
| ☐ | A percentage, with a denominator you had to choose | … |
| ☐ | A figure built from aggregated data | … |
Number ticked: … (fewer than four and this proposal comes back)
4 Language pair
| I will write the analysis in | R / Python / SAS |
| The assistant will translate it to | R / Python / SAS |
| Why this pair | one sentence |
| Assistant I plan to use | name it; models change, so this goes in the report with a date |
5 Question board
| Posted on | [date] |
| Questions already on the board using my dataset | … (or “none”) |
| How mine differs | … |
6 References
Codebook or help page, download URL and date, and anything else you have leaned on so far.