From One Finding to a Prototype: The 1-1-1-1-1 Rule
Where this fits
You know what a dashboard is and which pathway is required (dash01). This page is the translation step: how do you take one Week 7 finding and turn it into a small dashboard prototype? The answer is a rule with five ones.
The 1-1-1-1-1 rule
For your first dashboard prototype, commit to exactly:
- 1 audience — name a real person or role (e.g., “a public-health analyst preparing a briefing”).
- 1 question — phrased the way the audience would ask it.
- 1 control — one thing the audience can change (one dropdown, one filter, one editable value).
- 1 visualization — one chart that updates when the control changes.
- 1 limitation — one sentence about what the artifact does not support.
Beginners try to build a dashboard with three controls, four charts, and a tab system. That dashboard is harder to make, harder to test, and easier to break. The 1-1-1-1-1 rule keeps you honest.
A worked example using your Week 7 Table 1
Recall the Table 1 from eda03: mean BMI by income group across NHANES cycles. Here is the 1-1-1-1-1 translation:
| Field | Filled in |
|---|---|
| Audience | A public-health analyst preparing a short briefing on descriptive BMI patterns |
| Question | “Among adults age 20-80, how does mean BMI vary across NHANES cycles for one income group?” |
| Control | A single selected_income value the analyst can change |
| Visualization | A line plot of mean BMI by cycle for the selected income group |
| Limitation | The summary is descriptive and unweighted; it does not support causal or population claims |
That is exactly what the starter dashboard implements.
How to choose each field for your own project
Audience
Be specific. “The public” is too broad. “A first-year MPH student doing a course assignment” is specific. “A regional health officer planning a fall briefing” is specific. The more specific the audience, the easier every later choice becomes.
Question
Phrase it the way your audience would phrase it. Not “an exploration of the multivariate structure of NHANES BMI distributions” but “How does average BMI compare across income groups in this dataset?”. Plain language sharpens the rest of the design.
Control
One control. If you find yourself wanting two, pick the one your audience would change first. The second can wait for a later version.
Reasonable Quarto controls for this course:
- An editable R variable near the top of the
.qmd(the starter pattern — change the value, re-render). - A Quarto parameter declared in the YAML header (
params:block), updated at render time. - A
selectInput-style control inside a Shiny app — but only if you have approval for the Shiny pathway.
Visualization
One figure. Apply everything from Week 6: honest scale, descriptive title, weighted/unweighted disclosure, missing-value note, caption template. The dashboard does not relax those standards — if anything, it raises them, because an audience-facing artifact reaches more readers.
Limitation
One sentence. Use the templates from eda05 on what descriptive comparison cannot establish. Place the sentence where the audience will actually see it — at the bottom of the figure caption, or under the table, or beside the control.
What to write down before you build
Before you touch any code, write three lines in a note:
- My audience is: ______
- My audience would ask: ______
- The control I will offer is: ______
If you cannot finish those three lines, the dashboard is not ready to build yet. Go back to Week 7 and pick a finding you can actually frame.
A note on scope creep
Once you build your first prototype, you will be tempted to add a second control, a second figure, an “advanced settings” panel, or interactive tooltips. Resist. Every addition multiplies the surface area for misleading the audience. The A7 rubric does not reward extra controls; it rewards a clearly framed audience-facing artifact.
Where this goes next
dash03 is the working Quarto starter dashboard that implements the 1-1-1-1-1 worked example. You will adapt this file for A7.