Evaluating Source Credibility
Learn to judge a source's credibility by checking authorship, currency, purpose and bias, and corroboration — plus how to read laterally to investigate a source itself.
What you'll do in this lesson
A voice-first session with the Crimsora tutor on Evaluating Source Credibility, then targeted practice and FRQs — with the tutor adapting to where you get stuck.
What this lesson covers
You have a research question and a pile of possible sources. Now comes the harder part: deciding which ones deserve a place in your paper. A page can look polished, cite numbers, and still be built to sell you something — while a plain-looking government data table might be exactly what your argument needs.
This lesson gives you four questions to interrogate every source with: Who made it and what do they actually know? How current is it? Why does it exist? And who else confirms it? Then it adds the move professional fact-checkers use most: reading laterally — leaving the page to find out what other, independent sources say about this source. By the end you should be able to defend, in a sentence or two, why each source in your bibliography earned its spot.
This lesson gives you four questions to interrogate every source with: Who made it and what do they actually know? How current is it? Why does it exist? And who else confirms it? Then it adds the move professional fact-checkers use most: reading laterally — leaving the page to find out what other, independent sources say about this source. By the end you should be able to defend, in a sentence or two, why each source in your bibliography earned its spot.
Authorship and Expertise: Who Is Behind the Words?
Credibility starts with a name. Before you evaluate a claim, find out who is making it and what qualifies them to make it. Look for a byline, an "About" page, a publisher, or a sponsoring organization. If you cannot find any of these, that absence is itself information.
Expertise is domain-specific, not general. A cardiologist writing about heart disease is an expert; the same cardiologist writing about climate policy is an interested amateur. So ask three narrower questions: Does this person have training, credentials, or direct experience in this subject? Have they published elsewhere on it? Does an institution — a university, a hospital, a newspaper with an editorial process — stand behind the work?
Also separate the author from the publisher. An excellent researcher can be published on a site with no editorial standards, and a respected publisher can run a weak opinion column. Both layers matter.
Where students go wrong: treating a domain suffix as a credential. Anyone can register a .org, and .org signals only that a group set up a site, not that the group is neutral or expert. Similarly, a professional-looking design costs almost nothing to produce. Conversely, do not dismiss a source just because you have never heard of it — an unfamiliar name is a reason to investigate, not a verdict.
Eyewitnesses count too. For a research question about a local event, a firsthand account from a person who was there is a primary source with real value, even without formal credentials — as long as you label it as one account rather than as established fact.
Expertise is domain-specific, not general. A cardiologist writing about heart disease is an expert; the same cardiologist writing about climate policy is an interested amateur. So ask three narrower questions: Does this person have training, credentials, or direct experience in this subject? Have they published elsewhere on it? Does an institution — a university, a hospital, a newspaper with an editorial process — stand behind the work?
Also separate the author from the publisher. An excellent researcher can be published on a site with no editorial standards, and a respected publisher can run a weak opinion column. Both layers matter.
Where students go wrong: treating a domain suffix as a credential. Anyone can register a .org, and .org signals only that a group set up a site, not that the group is neutral or expert. Similarly, a professional-looking design costs almost nothing to produce. Conversely, do not dismiss a source just because you have never heard of it — an unfamiliar name is a reason to investigate, not a verdict.
Eyewitnesses count too. For a research question about a local event, a firsthand account from a person who was there is a primary source with real value, even without formal credentials — as long as you label it as one account rather than as established fact.
Currency: Is This Information Still True?
Currency means how recent a source is, and how much that recency matters for your specific question. There is no universal expiration date. A 1998 article on smartphone use in schools is useless; a 1998 close reading of a Shakespeare sonnet may still be excellent. The rule is that the faster the field changes, the tighter your currency requirement.
To judge currency you need a date, so hunt for one: publication date, "last updated" line, copyright year at the bottom, or the dates of the sources it cites. A page with no date and only decade-old references is quietly stale even if it looks new.
A common error is confusing an updated web page with updated content. Sites sometimes refresh a template or a copyright year while the article body stays untouched. Check whether the substance has changed — are the statistics, studies, and examples recent? Another error is the opposite: throwing out a source because it is old when your question is about the past. If you are arguing about how attitudes have shifted, an older source is evidence, not a liability.
| Type of question | How current the source needs to be |
|---|---|
| Medical, technological, scientific findings | Usually within about five years, or the most recent study available |
| Laws, policies, statistics, prices | The current version; superseded law is only useful as history |
| Ongoing news events | Days or hours; check for later corrections |
| Literary analysis, historical interpretation | Age is fine; landmark older criticism is often the strongest |
| Historical primary sources | Old is the point — an 1863 letter is valuable because it is from 1863 |
A common error is confusing an updated web page with updated content. Sites sometimes refresh a template or a copyright year while the article body stays untouched. Check whether the substance has changed — are the statistics, studies, and examples recent? Another error is the opposite: throwing out a source because it is old when your question is about the past. If you are arguing about how attitudes have shifted, an older source is evidence, not a liability.
Purpose and Bias: Why Does This Source Exist?
Every text was made for a reason: to inform, to persuade, to sell, to entertain, to advocate for a group, or to build a reputation. Identifying that purpose tells you how to use the source — and whether to use it at all.
Start with the money and the mission. Who funds this? A study on sugar and health funded by a beverage trade association is not automatically wrong, but the funding is a fact your reader deserves to know. An advocacy group states its position openly in its mission statement; that transparency is useful, because you can then read its evidence knowing which direction it leans.
Bias is not the same as lying, and it is not disqualifying. Almost everything has a perspective. What matters is whether the source's perspective causes it to distort: does it omit strong counterarguments, cherry-pick the friendliest statistic, use loaded language, or present opinion as established fact? Signals to notice include emotionally charged word choice, absolute claims with no cited evidence, sources that only ever cite themselves, and a total absence of any acknowledged opposing view.
Use biased sources deliberately rather than accidentally. An advocacy report is a fine source for what that side argues and for the evidence it assembles, as long as you attribute it: "According to a report from an organization that campaigns for later school start times..." That sentence is honest and actually strengthens your credibility with a reader.
Where students go wrong: labeling a source "biased" and stopping there, as if the word settled the matter. The stronger move is to name the purpose, name the likely slant, and then explain what you can and cannot responsibly draw from it.
Start with the money and the mission. Who funds this? A study on sugar and health funded by a beverage trade association is not automatically wrong, but the funding is a fact your reader deserves to know. An advocacy group states its position openly in its mission statement; that transparency is useful, because you can then read its evidence knowing which direction it leans.
Bias is not the same as lying, and it is not disqualifying. Almost everything has a perspective. What matters is whether the source's perspective causes it to distort: does it omit strong counterarguments, cherry-pick the friendliest statistic, use loaded language, or present opinion as established fact? Signals to notice include emotionally charged word choice, absolute claims with no cited evidence, sources that only ever cite themselves, and a total absence of any acknowledged opposing view.
Use biased sources deliberately rather than accidentally. An advocacy report is a fine source for what that side argues and for the evidence it assembles, as long as you attribute it: "According to a report from an organization that campaigns for later school start times..." That sentence is honest and actually strengthens your credibility with a reader.
Where students go wrong: labeling a source "biased" and stopping there, as if the word settled the matter. The stronger move is to name the purpose, name the likely slant, and then explain what you can and cannot responsibly draw from it.
Corroboration and Reading Laterally
The final test is whether anyone else, working independently, says the same thing. Corroboration means confirming a claim in a second and third source that did not get their information from the first one. Three websites all summarizing the same press release are not three sources; they are one source repeated. Trace claims back toward their origin — most credible articles will name the study, agency, or document they are drawing on, and you should be able to find it.
The most powerful habit here is lateral reading. Instead of staying on a page and scrutinizing its design, its "About" text, and its confident tone — which is what the page wants you to do — open new tabs and search for what other sources say about the site, the author, and the organization. Fact-checkers do this within seconds of landing on an unfamiliar page.
A practical routine: open the page, note the claim, then search the author's name, the organization's name, and the central claim itself. If reputable outlets describe the organization very differently from how it describes itself, believe the outside sources over the self-portrait.
Where students go wrong: counting agreement as corroboration when the agreeing pages all trace to the same original. Ask each time, "Did this source do its own work, or is it echoing?"
The most powerful habit here is lateral reading. Instead of staying on a page and scrutinizing its design, its "About" text, and its confident tone — which is what the page wants you to do — open new tabs and search for what other sources say about the site, the author, and the organization. Fact-checkers do this within seconds of landing on an unfamiliar page.
| Vertical reading (staying on the page) | Lateral reading (leaving the page) |
|---|---|
| Reads the site's own "About Us" | Searches the organization's name plus words like funding, criticism, or founder |
| Judges credibility by tone and design | Judges credibility by outside reporting and reference sources |
| Accepts the site's self-description | Compares self-description to independent descriptions |
| Checks whether the article cites sources | Clicks through to see whether those sources actually say that |
Where students go wrong: counting agreement as corroboration when the agreeing pages all trace to the same original. Ask each time, "Did this source do its own work, or is it echoing?"
Key terms
- Credibility.
- The degree to which a source can reasonably be trusted for a particular claim, based on its authorship, currency, purpose, and corroboration.
- Lateral reading.
- Evaluating a source by leaving it and searching for what independent sources say about its author, publisher, or funder, rather than judging it only by its own contents.
- Corroboration.
- Confirmation of a claim by a second or third source that arrived at the information independently, not by repeating the first source.
- Currency.
- How recent a source is, judged against how quickly information in that field changes.
- Bias.
- A consistent leaning toward a particular position or interest; not automatically disqualifying, but something a reader must know about and account for.
- Purpose.
- The reason a text was created — to inform, persuade, sell, advocate, or entertain — which shapes what it includes and leaves out.
- Primary source.
- A firsthand record produced by a participant or witness, such as a letter, interview, photograph, dataset, or original study.
- Domain suffix.
- The ending of a web address (.com, .org, .edu, .gov); it indicates registration category, not accuracy or neutrality.
Worked example
You are researching the question, "Should high schools start the school day later than 8:00 a.m.?" You find an article titled "The Sleep Crisis Stealing Your Teen's Future" on a site called RestRightNow. It has a clean design, a bar graph, and the line "Studies show teens who start school after 8:30 lose 40 percent less sleep." There is no author byline and no date. Evaluate this source.
Step 1 — Authorship. Look for a byline: none. Check the "About" page: it says RestRightNow is "dedicated to helping families rest better." No named staff, no credentials, no institutional backing. Nothing here establishes expertise in adolescent sleep science, so the authorship test is not satisfied on the page itself.
Step 2 — Currency. No publication or updated date appears. The footer copyright shows the current year, but that is a template, not evidence about the article body. Scan for internal date clues: the graph cites no year. Sleep and school-schedule research is active and policy-relevant, so an undated article fails the currency test — you cannot confirm it reflects current findings.
Step 3 — Purpose and bias. The headline uses fear language ("crisis," "stealing") and the site's name suggests a wellness brand. Scroll to the bottom: there are links to mattress and supplement products. That points to a commercial purpose. The article may still relay accurate research, but its incentive is to make you alarmed enough to buy something, which predicts selective use of evidence.
Step 4 — Corroboration and lateral reading. Open new tabs. Search "RestRightNow" — outside results describe it as a retail affiliate site, not a research organization. Now chase the claim itself: search for the statistic. You find that major pediatric and sleep-medicine organizations do recommend start times of 8:30 a.m. or later, and published studies report measurable gains in sleep duration — but no reputable source states the "40 percent less sleep" figure. The site invented or badly garbled a number.
Step 5 — Verdict and next move. Do not cite RestRightNow. Instead, cite the pediatric association's policy statement and the peer-reviewed study behind it — sources you found precisely because lateral reading pushed you upstream to the origin of the claim.
Step 2 — Currency. No publication or updated date appears. The footer copyright shows the current year, but that is a template, not evidence about the article body. Scan for internal date clues: the graph cites no year. Sleep and school-schedule research is active and policy-relevant, so an undated article fails the currency test — you cannot confirm it reflects current findings.
Step 3 — Purpose and bias. The headline uses fear language ("crisis," "stealing") and the site's name suggests a wellness brand. Scroll to the bottom: there are links to mattress and supplement products. That points to a commercial purpose. The article may still relay accurate research, but its incentive is to make you alarmed enough to buy something, which predicts selective use of evidence.
Step 4 — Corroboration and lateral reading. Open new tabs. Search "RestRightNow" — outside results describe it as a retail affiliate site, not a research organization. Now chase the claim itself: search for the statistic. You find that major pediatric and sleep-medicine organizations do recommend start times of 8:30 a.m. or later, and published studies report measurable gains in sleep duration — but no reputable source states the "40 percent less sleep" figure. The site invented or badly garbled a number.
Step 5 — Verdict and next move. Do not cite RestRightNow. Instead, cite the pediatric association's policy statement and the peer-reviewed study behind it — sources you found precisely because lateral reading pushed you upstream to the origin of the claim.
Practice questions
A student finds a report arguing against later school start times, published by an organization whose website states that it represents school bus contractors. What is the strongest way to handle this source?
- Discard it immediately, because an organization with a financial interest cannot be credible
- Cite it as neutral evidence, since it contains real data about transportation costs
- Use it while naming the organization and its interest, and check its cost figures against an independent source
- Use it only if it is fewer than five years old, since currency is the deciding factor
Answer: Use it while naming the organization and its interest, and check its cost figures against an independent source
Bias is a reason to read carefully and attribute openly, not an automatic disqualification. The organization has genuine expertise in transportation logistics, so its cost data may be the best available — but its stake in the outcome means the figures need corroboration from a source that does not profit from the answer. Discarding it outright throws away real evidence; presenting it as neutral hides information your reader needs; and currency alone cannot settle a question about purpose and interest.
Explain the difference between vertical reading and lateral reading, and give one specific example of a lateral-reading search you would run on an unfamiliar website called "The Center for Educational Progress."
Answer: Vertical reading means evaluating a source by staying on its own page — reading its About section, judging its design, tone, and citations. Lateral reading means opening new tabs and finding out what independent sources say about the site itself. A specific lateral search: type "Center for Educational Progress funding" or "Center for Educational Progress who founded" into a search engine and read what news outlets, reference works, or watchdog organizations report about the group, rather than accepting its self-description.
The key insight is that a page's self-presentation is written by the party being evaluated, so it is the least reliable place to learn whether that party is trustworthy. Any organization can call itself a nonpartisan center. Lateral reading works because outside coverage — reporting on its funders, its founder, its track record — was not written to make you trust the site. Strong answers name a concrete search string and explain why outside information outranks self-description.
A student writes: "My source is credible because it is a .org site with a lot of statistics and it was updated this year." Identify two weaknesses in this reasoning and state what the student should check instead.
Answer: First, .org indicates only a registration category; advocacy groups, businesses, and individuals can all register .org domains, so the suffix says nothing about expertise or neutrality. Second, "a lot of statistics" and "updated this year" describe surface features — statistics can be uncited, garbled, or cherry-picked, and a refreshed page or copyright year does not mean the article's content or data were revised. The student should instead identify the author and their expertise in this specific subject, find the publication or revision date of the article body, name the organization's purpose and funding, and confirm at least one central claim in an independent source by tracing it to its origin.
This question targets the two shortcuts students rely on most: domain suffixes and visual signals of rigor. Both are cheap to fake and easy to misread. The productive replacement is a set of questions about people, dates, motives, and independent confirmation — the same four tests applied in order, finished with a lateral search that leaves the page.
FAQ
- Is Wikipedia a credible source for a research paper?
- Treat it as a starting point rather than a citation. Wikipedia is often accurate on well-established topics and is genuinely useful for orienting yourself and for lateral reading about unfamiliar organizations. But it can be edited by anyone and it is a tertiary summary, so most teachers do not want it cited. The productive move is to scroll to the reference list at the bottom, follow the footnotes to the original studies, news reports, and documents, and cite those instead.
- How old is too old for a source?
- It depends entirely on your question. For technology, medicine, current law, or statistics, aim for the most recent reliable version — often within about five years. For literary criticism, historical interpretation, or philosophy, a source from decades ago can be a landmark. And when you are studying the past itself, old documents are exactly what you want. Ask: could the facts here have changed since publication? If yes, get something newer.
- What if every source I find says the same thing — is that corroboration?
- Only if the sources reached that conclusion independently. Ten articles that all summarize the same press release or all quote the same study are one piece of evidence repeated ten times. Trace each claim upstream to its origin. Real corroboration looks like separate research teams, separate datasets, or separate eyewitnesses arriving at similar findings.
- Can I ever use a source that is clearly biased?
- Yes, if you use it honestly. A biased source is strong evidence for what a particular side argues and often assembles useful data. Attribute it explicitly in your sentence — name the organization and its position — and verify any factual claim you borrow against a source that does not share its interest. What you should not do is present an advocacy document as if it were neutral reporting.
Learn this with a teacher, not a page
The Crimsora tutor teaches Evaluating Source Credibility live — explaining on a whiteboard, asking you questions, and adapting to where you get stuck.