Say you’re preparing a briefing for work and find a website you don’t know with an answer you could use. You read its explanation, then what it says about itself. You still don’t know who else has looked at its work.
Reading more from the same website may tell you what it claims. Deciding whether to rely on it also means looking beyond its own account. The useful question is a specific one: what reason do you have to trust this source on this topic?
This module is for working adults judging a source they don’t know, on a topic where you can look into the right expertise and evidence from elsewhere.
What you’ll get from this module
In this module
By the end of this module, you’ll have a clearer way to look into a source you don’t know and explain why you trust it on one topic. We’ll link teaching in college, experiments with adults, a report on expertise and an official correction in science with the decisions you face. You’ll choose a situation from daily life and finish a written check of a source. You’ll end with a reason to trust the source for now, or a clear account of what you still need to know.
Ask Jing
Become a Jingius.
What makes a source check useful?
A college study: learning to leave the page you started on.
Jessica E. Brodsky and colleagues’ 2021 study looked at lateral reading in university civics classes. Lateral reading means leaving the website you started on to look into the source, its claims or where they came from, using other information.
The researchers compared classes taught by the same teachers. Some classes were taught lateral reading; others followed the usual course. The teaching took six class sessions and three online assignments. Students practised looking into sources and explaining their judgements.
After the teaching, students answered a question about whether to trust a source. In the taught group, 14.2% of the 106 who answered both read laterally and reached the right conclusion. In the usual-course group, 1.2% of 84 did both. Missing answers and a video that wouldn’t play cut the numbers answering this question.
The top score needed both looking beyond the first page and reaching the right conclusion. Opening another page was only part of the task. Most students still didn’t reach that score, even in the taught group.
Why it matters: The classes were matched by teacher, not picked at random. The result shows promising gains on a set college task, after teaching and practice. Learning this skill takes more than hearing its name: you need practice judging what you find.
Read the college study of lateral-reading instruction.
An experiment with adults: a short introduction has a modest result.
Lisa Oswald and colleagues’ 2026 study tested a five-minute video on lateral reading with a German online panel of adults. People were picked at random to see the video, to get written instructions on searching or to be in a comparison group. They judged chosen websites and claims.
The video brought a small, immediate improvement in telling believable claims from unbelievable ones. Its effect on telling trustworthy sources from untrustworthy ones was unclear. Those are different judgements: judging one claim and judging the source behind it.
After about two weeks, the tests could no longer show a clear difference between the groups who got help and the comparison group. All groups had improved, and all had been told about the websites after the first round of testing. That pattern doesn’t show that all the learning was lost.
Why it matters: So a short introduction offers modest gains: a small, immediate benefit in judging claims, and an unclear one in judging sources. A demonstration can introduce the approach; practising it means looking at why the information you find deserves trust.
Read the adult experiment on lateral reading and search.
An experiment on searching: more information can back the wrong conclusion.
Kevin Aslett and colleagues’ study, published online in 2023, tested what happened when people were urged to search while judging recent news stories. In its first experiment, people on a US panel were picked at random to be urged to search or not.
Being urged to search raised the chance of rating a false or misleading story as true by 5.7 percentage points. This result came from 2,275 judgements of stories. On a separate measure, people rated how true each story was on a seven-point scale. That result didn’t show a clear effect.
The experiment tested being urged to search, not training in lateral reading focused on the source. The German trial with adults just described found small good effects from written instructions on searching. Together, these results show why one verdict about searching for everyone would mislead.
Why it matters: Finding another statement is different from finding a good reason to believe it. When two sources agree, you still need to know where their information comes from and whether someone qualified has looked at it on their own.
Read the experiment on searching to evaluate misinformation. Read the contrasting adult search experiment.
A report on ideas: expertise must fit the question.
Jonathan Osborne and colleagues’ 2022 report asks how people outside a field of science can judge whom to believe. It explains how to look for expertise in the right field and find out who has checked a claim.
The report points to qualifications or experience in the right field, and to being recognised by other experts in that field. It also asks readers to think about interests that could sway a judgement and how much is still uncertain. Expertise in one subject doesn’t show expertise in another.
This changes what you’re trying to find out. You may not be able to redo a technical analysis yourself, yet you can still look into whether people qualified in that field have checked it. Their reasons matter more than a general reputation for being impressive.
Why it matters: These are the report’s ideas about how to judge expertise. It isn’t an experiment showing that the checks settle every technical dispute. When the right experts truly disagree, you may still not know enough to choose between their conclusions. Seeing that gap is part of an honest judgement.
Read the report on science education and expertise.
An official account of a correction: a trusted process can change a result.
CERN’s account describes an unusual measurement of how fast neutrinos travelled, reported by the OPERA experiment in September 2011. The result was opened up for other scientists to examine.
In its June 8, 2012 update, CERN reported four experiments with results in line with the speed of light. It put the earlier oddity down to a faulty part in the experiment’s fibre optic timing system.
Technical checking changed the conclusion. The account gives you something to compare with your own decisions about trust: relying on qualified work can include accepting a correction to it.
Why it matters: You can keep the process apart from one result. Taking a correction seriously doesn’t mean treating all earlier expertise as worthless. It means going back to the specific reason you had for accepting the claim that changed.
Read the official account and its measurement updates.
What this means for you
Use a check of a source to explain why you would trust someone on one topic, and where that trust stops. You can rely on them for now while keeping the evidence to hand in case you need to think again.
Name the source and the topic
Start with who is responsible for the information and what you want to rely on them for. Treat what they say about themselves as a lead to follow up.
A general label such as “expert” leaves the important question half answered. The right expertise depends on the subject of the claim. Naming that subject keeps your conclusion tied to what you actually checked.
In the made-up briefing, write down the website and the one topic you need it for. Keep your judgement tied to the topic you checked. It doesn’t cover everything the website might publish.
Follow outside evidence to its origin
Leave the page you started on and look for information about the source’s expertise in that field, where its evidence comes from and who has checked it. Open useful results and look at their reasons.
Say two outside news pieces repeat the same source’s announcement. Their agreement still leads back to that one announcement. You’ve found the same news repeated, without showing that anyone checked the evidence separately.
Now imagine an outside account names experts in the right field who checked the work and explains their reasons. Now you have something specific to judge: who checked the work, what qualifies them to judge it and what they concluded.
Check the outside source too. Trace what it relies on and whether that support is separate from the first source’s own account. A result near the top of a search, or several matching pages, still leaves you needing to look at the evidence behind them.
Leave genuine gaps unresolved
If you can’t find outside evidence, write down what you couldn’t confirm. A lack of coverage doesn’t show that a source is dishonest. You may simply not have enough evidence to rely on it.
If the question needs a specialist’s judgement, look for evidence from the right experts or ask one. When qualified experts disagree and you can’t judge the dispute, leave the conclusion open.
This is a useful result because it shows what your next decision depends on. For example, “I need someone qualified in this field to check the claim” tells you what’s missing. A yes or no with nothing behind it would hide that gap.
Keep the reason, and revisit changes
When you have a reason to trust a source, write down the topic and that reason. Keep a note of the evidence if the source will matter again. Then you have a record to come back to when something changes.
Think again if the source moves beyond the expertise you checked. Do the same after an important contradiction, correction or retraction. Also think again when who is responsible, how they work or what they gain changes in a way that matters to your reason for trust.
Compare the new information with your first reason. Change the conclusion it affects when the evidence calls for it; write down any conflict you can’t settle. The CERN account shows why accepting a correction can fit with taking expert checking seriously.
For a source you use again and again, now and then look at a claim that matters to your decisions, and at its history of corrections. Ask whether your reason for trust still holds, rather than relying on a familiar name.
How about your daily life?
Where do you use information from sources you haven’t looked into? You might be preparing a briefing, reading an explanation you don’t know or going back to a website for guidance on one subject.
Write down situations where a clearer reason for trust would help. One suitable idea is enough.
| Situation | What I want to understand |
|---|---|
| ____________________ | ____________________ |
| ____________________ | ____________________ |
Choose one situation with a source and topic where you can look into the right expertise and evidence from elsewhere. If you find those aren’t there, naming the gap can be your result.
My chosen situation: ____________________
Carry this situation into the exercise. You can change your choice.
Check the basis for trusting one source
Start with your chosen situation. If you came straight here, choose a source you don’t know that you might rely on for one topic. Look for the right expertise and evidence from elsewhere; leave missing or disputed support open.
- Name the source and topic. Say who is responsible and what you want to rely on them for. Use what they say about themselves as a first lead.
Source, responsible person or organization, and topic: ____________________ - Investigate outside the source. Look for information about the right expertise, where the evidence comes from and whether people qualified in the field have checked it. Open useful results instead of treating their place in the list, or how many there are, as a verdict.
Outside evidence and where I found it: ____________________ - Assess that outside evidence. Trace it back to where it came from. Does it show that someone qualified checked the evidence on their own, or does it repeat the source’s own account? Explain what makes the outside source worth relying on.
If evidence is missing or the right experts disagree, leave the judgement open. Look for evidence from the right experts, or ask one, when a specialist’s judgement is needed. A lack of coverage on its own doesn’t show dishonesty.
What supports the judgment, or what remains unresolved: ____________________ - State what you can trust, and why. Finish one of these sentences: “I can provisionally rely on this source for this topic because…” or “I do not yet have a basis for reliance because…”.
My conclusion and evidence reference: ____________________
Keep this record if the source will matter again. Come back to it after a change in the right expertise, who is responsible, how they work or what they gain that affects your reason for trust. An important contradiction, correction or retraction is also a reason to think again.
Compare new evidence with your first reason and update the conclusion it affects. If you use the source often, now and then check a claim that matters to your decisions and its history of corrections. Leave conflicts open when you can’t settle them.
Ask Jing
Become a Jingius.
Your next step
You have finished when you can name the topic and explain your reason for trusting the source for now. If you can’t, name the missing support and the information or expertise you need next.
A note on the evidence
The teaching studies are about set judgements made with teaching or guidance. Their findings don’t show lasting benefits at work when people use the skill on their own. The adult follow-up matters most here: the comparison group improved too, after all groups had been given explanations.
The experiments on searching had different results in different settings. The report on expertise offers reasoning about whom to rely on, while the CERN account shows science correcting itself. Use each of these for its own purpose.
How we prepared this module
This module links research on teaching, experiments with adults, a report on ideas and an official account of a correction. A made-up briefing and a comparison of outside news pieces explain the decisions. The written exercise helps you look at one source and keep the reason for your conclusion.
Sources
- Jessica E. Brodsky, Patricia J. Brooks, Donna Scimeca, Ralitsa Todorova, Peter Galati, Michael Batson, Robert Grosso, Michael Matthews, Victor Miller and Michael Caulfield. (2021-03-31). Improving college students’ fact-checking strategies through lateral reading instruction in a general education civics course. Cognitive Research: Principles and Implications, 6, 23. Final article; Methods: Procedure, The DPI curriculum, Scoring of lateral reading problems; Results Table 5, sourcing row and footnote a.
- Lisa Oswald, Anastasia Kozyreva, Stefan M. Herzog, Pietro Leonardo Nickl and Ralph Hertwig. (2026-07-06 online; August 2026 issue). Boosting Media Literacy Using Lateral Reading and Online Search Interventions. Psychological Science, 37(8), 527–539. Final article; Method; Results: Lateral-reading and online-search treatment effects; Discussion.
- Kevin Aslett, Zeve Sanderson, William Godel, Nathaniel Persily, Jonathan Nagler and Joshua A. Tucker. (2023-12-20 online; 2024-01-18 issue). Online searches to evaluate misinformation can increase its perceived veracity. Nature, 625, 548–556. Final article; “Measuring the online search effect”, Study 1 design and results; Figure 1 caption; Discussion.
- Jonathan Osborne, Daniel Pimentel, Bruce Alberts, Douglas Allchin, Sarit Barzilai, Carl Bergstrom, Janet Coffey, Brian Donovan, Kari Kivinen, Anastasia Kozyreva and Sam Wineburg. (2022). Science Education in an Age of Misinformation. Stanford University collaborative report. Original report; competent-outsider discussion, printed pp. 4–6; Indicators of Expertise in Science, printed pp. 16–18.
- CERN. (2011-09-23; updates through 2012-06-08). OPERA experiment reports anomaly in flight time of neutrinos from CERN to Gran Sasso. Official press release and updates; original first three paragraphs; UPDATE 8 June 2012, first two paragraphs.
