Source Credibility Evaluation Protocol

What This Skill Does

Generates a structured lateral reading protocol for evaluating a specific type of source — teaching students to check what OTHER sources say about a source rather than analysing the source itself in isolation. This approach directly follows the research finding that professional fact-checkers outperform both students and university professors at source evaluation because they read laterally (opening new tabs to check who's behind a source) rather than vertically (reading the source itself more carefully for "clues" about reliability). The output includes a step-by-step protocol, a teacher modelling script, source-type-specific red and green flags, and a student checklist. AI is specifically valuable here because effective source evaluation requires both general principles (lateral reading, claim tracing) and source-type-specific knowledge (what makes a news article credible is different from what makes a scientific study credible) — a combination that is difficult to teach without domain expertise.

Evidence Foundation

Wineburg & McGrew (2017, 2019) conducted landmark studies showing that professional fact-checkers evaluate sources fundamentally differently from students and even university professors. Fact-checkers use "lateral reading" — instead of staying on a source and looking for clues about its reliability, they immediately open new tabs to check what other sources say about the source's author, publisher, and claims. Students and professors, by contrast, use "vertical reading" — they stay on the page and look for surface credibility markers (professional design, .org domain, author credentials listed on the page) that are easily manipulated. Lateral readers took an average of 93 seconds to reach a correct evaluation; vertical readers took over 5 minutes and were more often wrong. Breakstone et al. (2021) found that the vast majority of US high school students cannot reliably evaluate online sources — they are easily deceived by professional-looking design and on-page credentials. Caulfield (2019) operationalised the research into the SIFT method: Stop (pause before engaging), Investigate the source (who's behind it?), Find better coverage (what do other sources say?), Trace claims (find the original source). This framework makes lateral reading teachable. Hobbs (2010) established that digital and media literacy requires structured instruction — students do not develop source evaluation skills naturally through internet use.

Input Schema

The teacher must provide:

• Source type: What kind of source. e.g. "Website" / "News article" / "Social media post (Instagram infographic)" / "YouTube video" / "Wikipedia article" / "Scientific study abstract"
• Evaluation context: Why students are evaluating this. e.g. "Students are researching climate change for a Geography essay and need to find reliable sources" / "A student shared a TikTok video claiming X and I want the class to evaluate it"
• Student level: Year group. e.g. "Year 9"

Optional (injected by context engine if available):

• Specific source: Description of the actual source being evaluated