The Star Wars Map: A Whole Survey in One Chart

1,186 people, 14 characters, one chart. How Correspondence Analysis turns an unreadable contingency table into a map — with Jar Jar exiled to the corner. No code.

Dataset: the FiveThirtyEight Star Wars survey — 1,186 respondents rating 14 characters, ranking the films, and answering demographics. Public and CC-licensed. Download the CSV and follow along in Stratum — every figure below is real, computed from this file.

In 2014, FiveThirtyEight surveyed 1,186 people about Star Wars. Buried in it is a beautiful little problem: they asked each respondent how they felt about 14 characters — Han, Luke, Leia, Vader, Yoda, Jar Jar, and the rest — on a favorable / unfavorable / unfamiliar scale.

Cross-tabulate character against rating and you get an 84-cell wall of numbers. Run the classic test on it and you learn exactly one thing:

χ² = 4,900.7 (df = 65), p < 0.0001 — the ratings are not independent of the character.
Cramér's V = 0.29 — a solid, real association on a 0–1 scale.

Great — but a significance test won't tell you which characters are alike, who the outliers are, or the shape of the disagreement. For that you stop reading the table and start mapping it. The tool is Correspondence Analysis (CA) — the categorical cousin of PCA — and it's exactly what that wall of numbers was hiding.

An 84-cell contingency table of Star Wars character by favorability rating — Character × favorability: strongly associated (χ² = 4,900, V = 0.29) — and completely unreadable as a table.

Step 1 — Reshape, then run it

CA wants two categorical columns, not 14 side-by-side ones. So the one bit of wrangling: select the 14 character columns and use Data ▸ Stack Columns — fourteen wide columns become one tall, tidy pair, Character and Favorability. Then Analyze ▸ Correspondence Analysis, Rows = Character, Columns = Favorability. No knobs to turn.

The Correspondence Analysis report in Stratum, showing inertia per dimension — The CA report decomposes the χ² into dimensions — and the first one alone carries 76% of it.

The report breaks all that χ² into dimensions, ordered by how much of the association each explains:

Dimension	Inertia	% of total	Cumulative
Dim 1	0.3243	76.3%	76.3%
Dim 2	0.0792	18.6%	94.9%
Dim 3+	0.0216	5.1%	100%

Two dimensions capture 95% of everything in the 84-cell table. That's the whole promise of CA: a wall of numbers collapses to a picture you can read.

Step 2 — Read the map

Right-click the report → Drill Down to Correspondence Map. Two clouds of points — the characters and the rating levels — where proximity means association (a character near a rating got that rating more than chance predicts).

The Correspondence Map of Star Wars characters, with beloved heroes clustered and Jar Jar alone in the corner — The whole cast on one chart. Horizontal axis: adored → despised. Jar Jar is pushed farther out than anyone.

Dimension 1 (76%) is a single human thing: love. Left to right, adored to despised.

The beloved cluster: Han, Luke, Leia, Obi-Wan, Yoda, R2-D2, C-3PO land almost on top of each other, next to Very favorably. The original-trilogy heroes aren't just liked — the survey treats them as nearly interchangeable in how beloved they are (all within ≈0.05 of each other on Dim 1).
The far end: Palpatine and Darth Vader — the characters audiences love to hate.
The outlier: Jar Jar Binks, pushed farther from the pack than anyone — the single most distinctive opinion in the dataset. The math had one job, "find what's most unusual," and it pointed straight at Jar Jar.

Dimension 2 (19%) found fame on its own — separating villains people know from characters people don't. Lando, Padmé, and Boba Fett drift toward Unfamiliar: not hated, just not watched closely. Nobody told CA that fame was a variable; it recovered it anyway. That is the whole pitch — it hands you the structure you didn't know to look for.

Inertia. CA's word for total variation — it equals the table's χ² divided by the sample size. Each dimension explains a share of it, exactly like a PCA component explains a share of variance. "76% of the inertia" means Dim 1 reproduces 76% of the whole table's association.

Step 3 — The same trick finds your segments

This isn't a Star Wars trick — it's how you surface structure in any survey: customers, employees, voters, patients. So point it at the pettiest question in the file: "Who shot first — Han or Greedo?" against a single column, "Do you consider yourself a fan?"

	Fan	Not a fan
Han shot first	268	57
Greedo shot first	151	46
"I don't understand this question"	126	180

Correspondence map of who-shot-first against fan status, fans next to Han and non-fans next to don't-understand — Fans land next to "Han"; non-fans cluster next to "I don't understand this question." Cramér's V = 0.40.

Cramér's V = 0.40 — the strongest association anywhere in the survey. Same three options, two populations living in different universes, and one chart draws the border between them. (And the film that actually won? Not the one with the marketing — The Empire Strikes Back took 289 first-place votes, the clear favorite. The data has taste.)

When to reach for Correspondence Analysis

The rule of thumb is simple:

Your variables are…	Use
Numeric (measurements, scores)	Principal Components (PCA)
Categorical (survey answers, choices, roles, diagnoses)	Correspondence Analysis

When your data is categories, you can't run PCA — and most people give up and stare at a contingency table. CA is the tool you were missing: it takes the χ² you already trust and turns it into a map you can actually read. One reshape, four clicks, and an 84-cell table becomes a single chart a whole room understands.

That's the difference between a chart and an analysis.

The dataset is free and so is the trial. Download Stratum, load the Star Wars file, and map your own survey — then see which of your respondents live in different universes.