10 ways to explore the data. Each produces a different kind of insight.
"What assumptions does this NYT piece make about healthcare?
"Is this tech profile actually neutral, or does it embed a position?
Paste an article. See where it sits on all five axes.
The foundational analysis. The article is read through structured analytical methods that examine what it states, what it assumes, and what it leaves out. Observations are extracted and scored on all five axes. The output is a score card showing position and confidence per axis with paraphrased evidence.
"Where does Obama actually sit on ownership — based on what he did, not what he said?
"Does NPR's coverage match its public-service mission?
Understand a person, outlet, or institution across three bands.
Builds a three-band profile — what they said, what they proposed, what they did. Scored against the anchor library on all five axes. The gap between bands is one of the tool's most diagnostic findings. Works for politicians, authors, musicians, media outlets, institutions — the band labels adapt to the entity type.
"Is Harris really on the same side as Reagan on ownership?
"How does Allende compare to Pinochet across all five axes?
Pick two figures. See them side by side.
Select any two profiles in the library and compare their positions across all five axes. Overlaid spectrum bars show where they align and where they diverge. The comparison highlights which axes show the biggest gaps and what the evidence says about why.
"Did Chile really snap back left after Pinochet?"
"How has the NYT editorial board drifted on ownership since 2005?"
Track a country, outlet, or author over time.
Plot scores chronologically to see how positions shifted across years or decades. For countries: every head of state since 1900, plotted in sequence. For outlets: coverage drift over time. For authors: how a writer's framing evolved. Uses existing evidence file data — zero API cost for country trajectories.
"How did Muslim coverage change between 1996 and 2006?"
"Did gun control framing shift after Columbine — or just tilt and snap back?"
How did coverage change after an event?
Define a subject, a hinge event, a before period and an after period. The tool scores articles from both periods and measures the delta — not just THAT the discourse moved, but which lenses drove the change. Distinguishes between tilts (temporary, reverts to baseline) and shifts (permanent repositioning).
"Did Fox and NYT cover the Amazon union vote the same way?"
"How wide is the spread when 10 outlets cover the same Supreme Court decision?"
How did different outlets cover the same event?
Paste 3-5 articles about the same event from different outlets. All are scored and plotted on the spectrum. The spread — how wide the positions range, where the center of gravity sits, which outlets are outliers — is the finding.
"Was this 2003 article centrist for its time, or was it always center-right?"
"How much has the window actually shifted on ownership since the 1990s?"
Score an article against its era.
Every article is scored two ways: absolute position (era-independent) and relative-to-window (where it sat compared to its contemporaneous mainstream). A 2003 article can be unremarkable centrism for 2003 and center-right by 2026 standards. The gap between readings IS the measured Overton window shift.
"What does the Babylon Bee's audience find so obvious it's funny?"
"What does McSweeney's never mock — and what does that tell you?"
What does a satirical outlet's audience take for granted?
Score satirical content honestly — the surface text reveals the audience's shared assumptions. What they mock shows what they find absurd. What they never mock shows their sacred cows. The gap between the surface score and the audience's actual position is the comedy.
"What does the WSJ editorial board actually look like across 2,000 columns?"
"Where does mainstream healthcare coverage cluster on ownership?"
Aggregate thousands of articles to see the big picture.
Query the analyzed corpus by axis, score range, time period, outlet, author, or topic. Produces distribution curves, percentage breakdowns, and specific exhibits from across the dataset.
"Do I see the same things the tool sees?"
"Which lenses am I naturally good at — and which do I miss?"
Score it yourself and compare with the tool.
A guided walkthrough that takes you through the same observations the tool uses. Answer plain-English questions about what you see. Place your own scores. Then compare with the tool's analysis. Designed for classrooms — linkable from Google Classroom, printable, completable in one session.
Build your own analysis.
Design custom measurement dimensions, define your own rubric levels, create custom anchors, and run analyses against your own calibration library. For researchers, institutions, and enterprises with specialized measurement needs.