Skip to main content

Assignment 5: Visualization Audit

Purpose

This assignment turns a descriptive summary into an honest visual explanation. You will create a ggplot2 visualization, audit a flawed plot, revise your own design, and write a caption that states the statistic, data source, exclusions, and limitations.

Learning Objectives

  • Create a clear ggplot2 visualization from the NHANES classroom dataset.
  • Identify misleading scale, title, grouping, missingness, and caption choices.
  • Revise a plot so the visual design matches the evidence.
  • Explain the descriptive and unweighted limits of the figure.
  • Connect visualization readiness to the M1 project proposal.

Inputs

  • Dataset: examples/nhanes-equity/data/nhanes_equity_v6.csv
  • Worked example: weeks/week06-visualization/viz03-worked-example.qmd
  • Studio activity: weeks/week06-visualization/viz06-studio-visual-audit.qmd
  • Walkthrough and reference: weeks/week06-visualization/viz07-assignment-a5-and-reference.qmd
  • Flawed plot starter in your personal workspace: assignments/assignment05-visualization/flawed-plot-code.R
  • Visual-audit template in your personal workspace: assignments/assignment05-visualization/visual-audit-note-template.md
  • M1 connection: milestones/m1-proposal/README.md

Tasks

  1. Create or use the folder assignments/assignment05-visualization/submission/.
  2. Review the Week 6 worked example.
  3. Choose one descriptive comparison from the NHANES CSV snapshot.
  4. Draft an initial plot using R and ggplot2.
  5. Audit the flawed plot starter and your own draft for scale choices, labels, missing context, aggregation, and overclaiming.
  6. Revise your plot so it supports a descriptive interpretation.
  7. Export the corrected plot as corrected-plot.png.
  8. Write a caption that includes data source, statistic, exclusions, and limitation.
  9. Add a plotly preview [optional — include only if relevant to your project]: only if it helps you inspect the figure; the required submission is still the static plot.
  10. Write a short M1 link note naming how the visualization standard affects your project dataset, audience, or feasibility.

AI-Use Expectations

AI may help draft ggplot2 code, suggest labels, or identify possible visual problems. You must verify the code against real column names, check that the scale does not exaggerate the result, and remove causal or population-level claims that the plot cannot support. Include ai-use-note.md.

Reproducibility Requirements

  • The Quarto file must render in GitHub Codespaces.
  • The dataset path must be relative.
  • The plot image must be exported by code or a documented render step.
  • Captions must state that the analysis is descriptive and unweighted when using the classroom NHANES snapshot.

What To Submit

Place all submission files in:

assignments/assignment05-visualization/submission/

Submit these files:

  • visualization-audit.qmd [required]: rerunnable Quarto file that loads the CSV, creates the final plot, and explains the revision.
  • visualization-audit.html [required]: rendered output.
  • corrected-plot.png [required]: exported final plot.
  • visual-audit-note.md [required]: at least three issues checked or corrected.
  • caption.md [required] [brief — 1-3 focused bullets or sentences are enough]: final figure caption.
  • m1-link-note.md [required] [brief — 1-3 focused bullets or sentences are enough]: two or three sentences connecting the visualization standard to M1.
  • ai-use-note.md [required]: what AI helped with and how you checked it, or a sentence saying AI was not used.

Submission Route

Submit your GitHub repository link on Canvas after committing and syncing the required files.

Grading Checklist

Canvas applies a 7-point rubric. This assignment is part of the best-4-of-6 set for Assignments 4-9:

Criterion Points
Plot is generated by rerunnable ggplot2 code with relative paths 1.5
Visual audit identifies at least three meaningful issues 1
Corrected plot improves scale, labeling, grouping, or caption clarity 1.5
Caption states statistic, source, exclusions, and limitations 1
Interpretation stays descriptive and non-causal 0.75
AI-use note and M1 link note are specific 0.75
Required files are committed and synced 0.5

Definition of Done

You are done when a peer can render visualization-audit.qmd, see the corrected plot, understand the main comparison from the caption, and identify how you checked the plot for misleading design.