Capstone Visualization Project: End-to-End Build and Peer Defense

Capstone Visualization Project: End-to-End Build and Peer Defense

This lesson is the culminating practice experience for the Data Visualization course. You will bring together every skill developed across the term—data cleaning, chart selection, accessibility design, and narrative structure—into a single, defensible portfolio piece. The result should demonstrate to any employer or credentialing examiner that you can take raw data from source to story without guidance.

What the Finished Portfolio Piece Contains

Your submission has four components that work together:

1. Static charts file – at least two publication-quality charts produced in Python (Matplotlib, Seaborn, or Plotly static export) or R (ggplot2). These must be saved at sufficient resolution (300 dpi is the professional standard for print; 96 dpi minimum for screen) and exported in a lossless or high-quality format such as PNG or PDF.

2. Tableau Public interactive dashboard – a published, publicly accessible dashboard with at least two coordinated views, a title, a data source citation, and working filters or tooltips.

3. Data story deck – a slide presentation (six to ten slides) that opens with a clear question or problem, walks through your key findings using your charts as evidence, and closes with a specific actionable insight. Slides are not a gallery of visuals—each slide carries one argument.

4. Design journal – one to two pages documenting your dataset source, cleaning steps, chart-type rationale, accessibility decisions, and misleading patterns you deliberately avoided.

All four must be submitted before the peer defense session begins.

Design Decisions You Must Be Able to Defend

Every choice you made—color palette, chart type, axis scale, level of aggregation—is fair game during your peer defense. Prepare to explain:

• Why this chart type? Bar charts compare discrete categories; line charts show change over a continuous variable, typically time; scatter plots reveal relationships between two quantitative variables; maps encode geographic distribution. Choosing the wrong type misleads readers even when the underlying numbers are correct.
• Why this color palette? Use a colorblind-accessible palette (such as the ColorBrewer diverging or qualitative schemes, or Tableau's color-blind palette) whenever categories must be distinguished. Avoid red-green combinations as the primary encoding. Sequential palettes work for ordered or continuous data; diverging palettes work when a meaningful midpoint exists.
• Why this axis range? Bar charts must start at zero because the eye reads bar length as proportional to value. Line charts and scatter plots may use a non-zero baseline when the range of variation is the message and that choice is labeled. Truncating a bar-chart axis to exaggerate small differences is a misleading practice you must document having avoided.
• Why this level of aggregation? Showing every individual data point can obscure trends; over-aggregating can hide important variation or Simpson's Paradox effects. Name the aggregation function you used (mean, median, sum, count) and justify it.

Accessibility Is Non-Negotiable

Accessibility decisions count for up to 25 points on the rubric. Apply these practices to every deliverable:

• Add descriptive alt text to every static chart if the file will be embedded in a web page or slide deck. Alt text states what the chart shows in one sentence, not just its title.
• Ensure sufficient contrast between text, marks, and backgrounds. The Web Content Accessibility Guidelines (WCAG) specify a minimum contrast ratio of 4.5:1 for normal text. Free tools such as WebAIM's Contrast Checker can verify this.
• Use redundant encoding: pair color with shape, pattern, or direct labels so that color alone never carries the only meaning.
• In Tableau, add tooltips that provide the exact values underlying each mark so that users who cannot distinguish colors can still read precise data.

The Peer Defense Format

Your group of four takes turns at the front. You have exactly 10 minutes to present your Tableau dashboard and one static chart. Walk your peers through the story, not the tool. Explain what question you started with, what the data showed, and what a decision-maker should do differently because of it.

After your 10 minutes, your peers have five minutes of structured questions drawn from the critique card. Critique-card questions target design rationale, accessibility choices, and potential misleading elements—not personal style preferences. Respond directly and concisely. If a peer identifies a genuine flaw, acknowledge it and state how you would fix it; defensiveness under critique is not professionalism.

The instructor scores each presenter independently using the four-criterion rubric. Peer questions do not determine your grade, but your responses to them inform the instructor's storytelling and design-principles scores.

Avoiding Common Last-Minute Errors

• Publish to Tableau Public before the session, not during it. Confirm the link works from a browser where you are not logged in.
• Save static chart files with descriptive file names and include the chart title inside the image, not only in the file name.
• Proofread axis labels, legend labels, and slide text. Misspelled labels and unlabeled axes are automatic deductions under the technical accuracy criterion.
• Cite your dataset on every deliverable. Include the source name, URL or accession number, and the date you accessed it. This is standard professional and academic practice.

Connecting to Your Credentials

The skills assessed in this capstone map directly to content tested in the Google Data Analytics Certificate (data cleaning documentation and chart selection), the Microsoft PL-300 exam (interactive dashboard construction and DAX-driven visuals in Power BI, conceptually parallel to Tableau), the IBM Data Science Professional Certificate (Python-based visualization and analytical storytelling), and CompTIA Data+ (data visualization best practices and communication of analytical findings). A strong capstone portfolio piece is also a concrete artifact you can show to employers before certification results arrive.