Week 9 — Tuesday

Finishing Your EDA & Connecting to the Final Project

STAT 80B - Data Visualization

Where Are We?

The home stretch

This week:

  • 📋 EDA due Friday at midnight
  • 🔬 Lab 5 on Thursday

Next week:

  • 🎤 Project Presentations (in class)
  • 📊 EDA submitted and done

Important

No extensions. Submit what you have by Friday.

Part 1: What is the EDA?

EDA in one sentence

The EDA is you exploring your data out loud — trying things, noticing patterns, and deciding which visualizations are worth polishing.

It does not need to be perfect. It needs to be honest and thorough.

What the EDA needs (checklist)

Section What to write Length
Introduction Dataset + updated questions 1 page
Visualizations 10–15 charts with short captions 5–8 pages
Findings What you discovered 1–2 pages
Final selection Which 5–7 charts go in the report 1 page
Reflection What you’d do differently 0.5–1 page

Total: 8–12 pages.

The visualizations: what counts

You need at least:

  • 5 different chart types
  • 2 multi-panel or small multiple figures
  • 1 chart showing distributions or uncertainty

Tip

Don’t know if you have enough variety? Look at your 10–15 charts and count the types. If you have 8 scatter plots and nothing else, add a bar chart, a box plot, or a histogram.

Writing captions (keep it simple)

Each visualization needs 2–4 sentences. Just answer:

  1. What does this chart show?
  2. What pattern or insight do you see?
  3. Does it help answer one of your research questions?

Example:

“This scatter plot shows the relationship between neighborhood income and average rent. Higher-income neighborhoods tend to have significantly higher rents, with a few outliers in the downtown area. This addresses research question 2.”

Part 2: EDA → Final Project

Think of the EDA as a draft

EDA (due Friday)          →    Final Report (Finals Week)
─────────────────────────────────────────────────────────
10–15 exploratory charts  →    5–7 polished charts
Short captions            →    Full captions (3–5 sentences)
Raw findings              →    Coherent visual story
"I tried this..."         →    "I chose this because..."

The EDA feeds directly into the final report. You are not starting over — you are selecting the best work and polishing it.

The Final Selection section is key

In your EDA, you have a section called “Final Visualization Selection.”

This is where you say: “Out of everything I tried, these are the 5–7 charts I’ll polish for the final report.”

For each one, write:

  • Which chart it is
  • Why you chose it (most insightful? clearest? answers a key question?)
  • What you plan to improve (colors, labels, title, layout)

Note

This section makes your final report much easier to write. Invest time here.

What “polishing” looks like

EDA version Final report version
Default colors Colorblind-safe palette
No title Descriptive title
Short axis labels Clear, full labels with units
Small font Readable at print size
1 chart per question Narrative connecting all charts

The design justification section (Final Report)

In the final report you’ll explain your choices. Your EDA is where you make those choices.

While working on your EDA, write down:

  • Why did you use this chart type?
  • Did you try a different chart first? What happened?
  • Why did you keep or drop certain variables?

These notes = your design justification section later.

Part 3: Work Session

The next 45 minutes

Work on your EDA with your partner. Use this time to:

  1. Audit your charts — do you have 10–15? Enough variety?
  2. Write or improve captions — 2–4 sentences each
  3. Write your Findings section — what did you actually discover?
  4. Fill in your Final Selection — pick your best 5–7 charts

We will circulate and answer questions.

🛠️ Activity: Audit your EDA (5 min)

Fill out this quick checklist with your partner:

Done?
10–15 visualizations ☐
5+ different chart types ☐
2+ multi-panel figures ☐
1+ distribution/uncertainty chart ☐
All charts have captions ☐
Findings section drafted ☐
Final selection identified ☐

Anything missing → work on it now.

🛠️ Activity: Final Selection (10 min)

With your partner, look at all your charts and answer:

  1. Which 3 charts best answer your research questions?
  2. Which chart shows the most surprising or interesting pattern?
  3. Which chart do you think looks the most polished already?
  4. Which 1–2 charts could you drop without losing important information?

Your final 5–7 should come from answers 1–3. Drop 4.

🛠️ Activity: Sketch your story (10 min)

Your final report needs a narrative flow — the charts should tell a story together.

On paper (or a blank slide), arrange your 5–7 selected charts in order and write one sentence connecting each to the next:

“First we show X… then we zoom into Y… this reveals Z…”

This will become the structure of your final report introduction.

Before Thursday

What to have done by Thursday

On Thursday: Lab 5 on design critique — you’ll practice evaluating and improving visualizations, which directly prepares you to polish your own charts for the final report.

Questions?

Ed Discussion → best for quick questions
Office hours → Tue & Thu 3:05–3:40 in the classroom
EDA due: Friday at midnight, no extensions