How Ad Fontes Media Rates a Source

Goal: Explain how Ad Fontes Media (ADM) rates a source of news.

In Brief

1. Get multiple samples from a source that span the source’s types of content.
2. Rate each sample on two scales: reliability and bias.
   1. Use three reviewers for each article: conservative, middle, liberal.
3. Create an average rating for the source that is based on all the samples analyzed to date.
4. Continually rate more samples and update the source rating.

More Detail

Content Sampling

ADM doesn’t rate a news source based on one article or impression. Instead, they:

• Select a sample of articles, videos, or podcast episodes from the outlet.

• Aim for a balanced and representative sample that includes both news and opinion pieces if applicable.

Multiple Analyst Ratings.

Each content item {article, editorial, cartoon, video, podcast, meme, … } is reviewed by at least three analysts:

• One left-leaning
• One right-leaning
• One center-leaning

This helps reduce individual bias in the evaluation.

Dimensions (Reliability and Bias )

Rating Dimensions: Each analyst scores the content item on two main scales:

1. Reliability (0–64 points): Assesses how trustworthy and informative the content is:
   • Is it original reporting?
   • Are the facts well-sourced?
   • Does it include misleading or sensational information?
2. Bias (-42 to +42 scale): Assesses political bias, including:
   • Language/tone
   • Story selection
   • Framing
   • Placement of opinions or political slants
   • Negative = left-leaning
   • Positive = right-leaning
   • Near zero = balanced/center

Averaging the Scores

The individual analyst scores are averaged for each piece, then across all sampled content from the source.

This gives:

• A Reliability Score
• A Bias Score

These are plotted on the Media Bias Chart.

Continuous Updates

ADM continually adds new articles to update the rating of a source over time. This allows for changes in editorial standards or practices to be reflected.

Summary

AFM creates a rating for a news source by analyzing a representative sample of its content using a multi-analyst team with varied political leanings, scoring each item for reliability and bias, and averaging those scores to produce a position on the Media Bias Chart.