A practical way to look beyond a persuasive page and decide what the available evidence actually supports
By Arbab Naseebullah Kasi, Chief Executive Officer - CEO, Feel Worldwide Foundation Inc.
A moment of comparison before passing information along.
A striking number appears in a post. The layout looks professional, the wording sounds confident and someone you respect has already shared it. Passing it along would take a moment. Understanding where it came from may take a little longer.
That pause is worthwhile. When we share a claim, we give it another route to people who may use it to make a decision. They may reasonably assume that we have looked beyond the headline, especially when we share it through a professional or organisational account.
The task does not require becoming an expert in every subject. It begins with a more manageable question: what would justify trusting this particular claim enough to repeat it? Sometimes a few checks provide a clear answer. Sometimes they reveal that the evidence is narrower than the post suggests. And sometimes the responsible choice is to leave it unshared.
A convincing page is only the starting point
A website can tell us a great deal about how it wants to be seen. Its name, logo and introductory language may suggest expertise. Those features deserve less weight than evidence about who produced the information and how the claim was established.
In a study published in 2019, Sam Wineburg and Sarah McGrew observed 45 experienced internet users: 10 professional fact checkers, 10 historians with doctorates and 25 Stanford undergraduates. Participants evaluated websites while explaining their thinking. The fact checkers frequently left the initial site and opened other tabs to investigate it. They reached better-supported judgments more quickly than the other groups, who often relied on features such as official-looking logos. This was a small, selected sample completing particular tasks, not a population survey or a guarantee that a browsing technique always produces the right answer. The original research abstract describes the study and its conclusions.
The practical idea is called lateral reading: use the wider web to investigate a source instead of asking that source to supply all the evidence of its own reliability. You can return to read closely once you have a better sense of what you are reading.
Identify the claim you are checking
A long post may contain a factual statement, an interpretation and a recommendation. Separate them before searching. Otherwise, you can spend time confirming something that was never the important point.
For example, imagine an illustrative post saying that a local training programme doubled its participation and therefore proves that its teaching method is the best available. The participation figure and the claim about teaching quality require different evidence. Even if the number is accurate, it does not by itself establish why participation changed or how much participants learned.
This is a hypothetical example, not an account of an actual programme. Its purpose is to show how easily one supported fact can carry a much larger conclusion beside it.
Write the central claim in a plain sentence. Include the population, place and time if they matter. “Participation increased in one programme during one year” is a more precise thing to investigate than “this approach works.” Precision makes the search smaller and the eventual explanation more honest.
Look beyond what the publisher says about itself
Search for the organisation or author in a separate tab. Look for reliable information about their relevant work, expertise and relationship to the subject. A self-description can help identify names and affiliations, but it should not be the only basis for deciding what those affiliations mean.
The Digital Inquiry Group provides a public example of teaching this practice. Its Intro to Lateral Reading lesson, reviewed in October 2026, introduces investigating an unfamiliar source through other trusted websites. The lesson uses teacher modelling followed by student practice. It is an educational resource, not an automated certificate of a website's trustworthiness.
When investigating a publisher, avoid reducing the result to a simple badge. A source may be authoritative about its own opening hours and less useful for judging the effectiveness of a product it sells. Relevant expertise matters more than an impressive title unrelated to the claim.
A financial or organisational interest is useful context, but it does not automatically make every statement false. Look at the evidence and how the interest is disclosed. Likewise, finding no criticism in a quick search is not proof that a source is reliable. It may simply be unfamiliar or poorly documented.
Follow the claim towards its evidence
Look for the report, dataset, official announcement or full interview on which the statement depends. Follow the actual reference rather than assuming that the presence of a hyperlink means it supports the sentence beside it.
Check whether the original material says what the post says. A report might describe respondents to a voluntary survey while the post refers to all workers. An announcement might describe a planned service while the summary says it is already available. A quotation might be accurate but omit the sentence that qualifies it.
For a numerical claim, ask what was counted and what the comparison is. In the hypothetical training example, an increase from 10 to 20 participants is a doubling, as is an increase from 1,000 to 2,000. The percentage alone does not convey the scale. It also leaves questions about attendance definitions, reporting periods and whether the same people were counted more than once.
You do not need to invent an alternative explanation. You need to identify what the evidence can and cannot establish. If the original material is unavailable, say that you could not verify the claim from its underlying source rather than treating a chain of summaries as equivalent evidence.
Check the date and the setting
A true statement can become misleading when it is detached from its circumstances. An old closure notice may be shared as if it describes today. A result from one region may be presented as a worldwide fact. A temporary offer may continue circulating after its conditions change.
Look for the publication date, the date of the event or data and any later correction or update. These may be different. A newly published article can discuss information collected several years earlier, and a recently shared screenshot may contain an old announcement.
The appropriate level of freshness depends on the question. A historical account does not become useless because it is old. Instructions for a service that changes frequently need a current check. Decide what you need the information to do before deciding whether its age is a problem.
Images deserve context too. If the date or location attached to an image is central to the claim, seek its original publication and caption. A visually plausible image does not establish where or when an event occurred. If that context remains uncertain, avoid presenting the image as proof.
Check whether apparent agreement is independent
Finding the same sentence on several websites can feel reassuring. Before treating that as corroboration, examine where each version obtained it. Several articles may repeat the same press release, report or social post.
Those repetitions can help locate the original material, but they do not necessarily provide additional evidence. Ask whether another source has independently checked the point, supplied relevant records or explained a competing interpretation using evidence you can examine.
Do not seek a contradictory opinion merely to make the discussion look balanced. A well-supported finding does not require an equally prominent unsupported objection. The useful distinction is between disagreement that identifies a real evidential problem and disagreement that simply repeats a preference.
For specialist questions, recognise when your own review reaches its limit. You may be able to locate the relevant research without being qualified to evaluate every method or application. Use appropriate expert analysis and keep high-stakes decisions separate from casual online sharing.
Practise the process without promising certainty
There is evidence that instruction can help people evaluate online information. A district-wide study reported by the Digital Inquiry Group in April 2022 tested classroom teaching of lateral reading. Its cluster-randomized analysis found improvement in students' ability to judge digital credibility in the experimental classrooms. The research summary describes the intervention. It does not establish that a single reminder will improve every adult's sharing decisions or prevent every mistake.
For personal practice, choose one claim that genuinely interests you and work through the questions carefully. Notice which step changes your understanding. Perhaps the source is credible but the headline overstates the finding. Perhaps the underlying evidence is clear and you can share it with better context.
Keep the effort proportionate. Not every ordinary conversation needs a research dossier. A claim that could influence money, reputation or someone's access to an opportunity deserves more care than a harmless preference. If you cannot give a consequential claim the attention it needs, you can refrain from amplifying it.
Share what you know without upgrading it
When the evidence supports sharing, preserve its limits in your wording. Identify who reported the information, the period or setting and any important qualification. Link to the strongest accessible source you actually checked rather than only to the most dramatic summary.
Avoid turning “the organisation reports” into “research proves.” Do not describe a preliminary result as settled or a local example as a universal rule. If you add your interpretation, make the distinction visible so readers can evaluate it separately.
In a team, a brief source note can help the next person. Record the claim, the original link, the date checked and what remains uncertain. Keep the note focused. The aim is to make the reasoning inspectable, not to bury an uncertain statement under a large pile of links.
If you later discover an error in something you shared, correct it where the original audience can reasonably see the correction. Explain the specific change and provide the better evidence. Quietly becoming more careful in future does not tell earlier readers that the information they received needs updating.
We can remain curious without treating every confident post as trustworthy, and careful without assuming everything is false. A useful pause gives us space to discover what is known before asking someone else to believe it.
What is one online claim you would want to trace back to its original evidence before sharing it again?
