E-Prime Bias Detector highlights the linguistic patterns that shape how you interpret information — opinion disguised as fact, vague sourcing, emotional manipulation — alongside markers of quality writing. The goal is sharper reading, not prescribed conclusions.
View on GitHubThe extension runs on any web page, highlighting patterns in real time. Here's what a typical analysis looks like:
Experts say the new policy is clearly a step in the right direction. Everyone agrees that change was long overdue.
According to Dr. Chen et al. (2024), the data suggests a possible correlation, though important limitations include the sample size and geographic scope.
Click any highlight to see contextual guidance — not just what was detected, but how to evaluate whether it matters:
Absolute terms that rarely reflect reality accurately
Each category targets a specific way language can mislead or demonstrate integrity. The extension doesn't assume any pattern is inherently bad — it makes it visible so you can decide. See the full pattern reference for every word and phrase the extension detects.
Subjective language and evaluative terms like "obviously," "terrible," "brilliant" — judgment embedded in description.
E-Prime violations that can create false equivalencies. "This policy is a failure" collapses complex evaluation into identity.
Universal claims like "everyone," "always," "never" — sweeping statements that are almost never literally true.
Constructions that obscure who is responsible. "Mistakes were made" hides who made them.
Vague attributions like "experts say," "studies show," "people are saying" — authority without accountability.
Hidden assumptions smuggled into statements as established fact. The embedded premise goes unquestioned.
Words that shape how likely we perceive events to be, often misaligning perception with statistical reality.
Militaristic language applied to non-military topics. Framing everything as a fight constrains how we think about solutions.
Language that downplays significance: "merely," "just," "only." Distorts scale by making things seem smaller than they are.
Exaggeration and hyperbole: "absolutely," "the greatest," "unprecedented." Emphasis substituted for evidence.
Presenting two positions as equally valid when evidence strongly favors one. Creates artificial equivalence.
Softening language that obscures harsh realities. "Collateral damage" for civilian deaths, "downsizing" for layoffs.
Language designed to bypass rational evaluation and trigger emotional responses instead of analysis.
Phrases that undermine confidence in documented facts: "that never happened," "concerns are overblown."
Presenting complex situations as binary choices: "you're either with us or against us." Reality is rarely that simple.
Bias detection alone creates cynicism. The excellence system shows what responsible, intellectually honest writing actually looks like — so you can recognize and reward it.
Named sources, dates, publication venues, sample sizes. Verifiable claims you can actually check.
Acknowledging complexity rather than pretending certainty. "The evidence suggests" rather than "this proves."
Being upfront about limitations, perspective, and uncertainty. Saying "in my opinion" when it is one.
Building on others' ideas, acknowledging valid counterpoints, inviting dialogue rather than shutting it down.
Specific data, methodology references, statistical context. Claims grounded in verifiable information.
Pre-built extensions are included in the repo — no build step required. Just clone and load.
Or download it as a ZIP from GitHub.
git clone https://github.com/shadnbob/e-primer.git
Chrome: Go to chrome://extensions, enable Developer mode, click "Load unpacked" and select the dist/ folder.
Firefox: Go to about:debugging, click "Load Temporary Add-on" and select dist-firefox/manifest.json.
Only needed if you want to modify the extension.
cd e-primer && npm install
npm run build # Chrome
npm run build:firefox # Firefox