Source Evaluation

Overview

You are going to learn how to tell the difference between information you can trust and information you cannot. This is not a lecture about "fake news." It is a hands-on skill — like learning to tell fresh fish from spoiled fish by its smell, its texture, and its eyes. You are going to develop a nose for unreliable information, and you are going to build a tool that helps you use it.

Right now, you are swimming in more information than any human being in history has ever encountered. Your grandparents got their news from a handful of newspapers and three television channels. You get yours from an ocean of websites, social media feeds, YouTube videos, podcasts, and group chats. Some of that information is carefully researched and verified. Some of it is wrong by accident. Some of it is wrong on purpose — designed to make you angry, scared, or confused so that someone can profit from your attention.

The problem is not that bad information exists. The problem is that bad information often looks exactly like good information. A slick website with professional graphics can be completely fabricated. A social media post shared by ten thousand people can be entirely false. A book published by a real publisher can contain serious errors. You cannot judge information by how it looks. You need a system.

By the end of this activity, you will have built a Reliability Scorecard — a personal tool for evaluating any source of information you encounter. You will have tested it on real sources and proven that it works. And you will have developed instincts that no algorithm can replicate: the ability to pause before believing something and ask, "How do I know this is true?"

The Deliverable

A completed Reliability Scorecard template (designed by you, tested on at least eight real sources), a scored evaluation folder containing your marked-up sources with ratings and written justifications, and a one-page "Red Flags" reference sheet listing the warning signs of unreliable information that you discovered through your own analysis.

Before You Start

Source Collection (Parent and Child Together)

Before Session 1, you and a parent need to gather 8-10 printed sources. This is the raw material for the entire activity. The collection must include a mix — some reliable, some unreliable, some in between. Do not tell the child which is which. The whole point is for them to figure it out.

Here is what to collect. Print each source so the child can mark it up with a pen.

Include at least one of each:

• A news article from a major newspaper (e.g., The New York Times, The Wall Street Journal, your local paper)
• A news article from a less well-known or clearly biased website
• A Wikipedia article on any topic
• A social media post making a factual claim (screenshot it and print it)
• A page from a .gov or .edu website
• An article from a blog or personal website
• A page from a product advertisement disguised as information (common in health, supplements, and technology)
• A passage from a nonfiction book (photocopy a relevant page)

Optional but excellent additions:

• A satirical article (from The Onion or similar) — this tests whether the child can recognize humor disguised as news
• An AI-generated article — easy to produce, and an increasingly important category to recognize
• A scientific press release alongside the actual study abstract it describes — this shows how information changes as it moves from source to summary

Label each printed source with a number (Source 1, Source 2, etc.) but do not add any other identifying marks.

Session 1: The Five Checks (60 minutes)

What Makes a Source Reliable?

Open your notebook to a fresh page. Write this heading: "The Five Checks." These are the questions you will ask about every source you evaluate. You did not invent them — journalists, historians, scientists, and librarians have been using versions of these for centuries. But you are going to learn them by doing, not by memorizing.

Check 1: Who wrote this?

The most basic question. A source is more trustworthy when you can identify the author and verify that they have relevant knowledge or credentials. An article about climate science written by a climate scientist is more credible than the same topic covered by someone with no scientific background. This does not mean non-experts are always wrong — but it means you should look harder at their evidence.

What to look for: Author name and bio. Professional affiliation. Other work by the same author. If no author is listed at all, that is a yellow flag — not an automatic disqualification, but a reason to be more cautious.

Check 2: What is the evidence?

Reliable sources show their work. They cite studies, quote experts, reference data, and link to original sources. Unreliable sources make claims without support, or support them with vague phrases like "studies show" or "experts say" without naming the studies or experts.

What to look for: Specific citations. Named sources. Links to original data. Quotations with attribution. Numbers with context (percentages should tell you "of what").

Check 3: Who benefits?

Every source exists for a reason. A newspaper article exists to inform (and to sell newspapers). A corporate press release exists to make the company look good. A nonprofit's report exists to support their mission. None of these motivations automatically makes the source wrong — but you need to know the motivation so you can account for the bias.

What to look for: Who published this? Who funded the research? Is the source trying to sell you something? Is it trying to make you angry or afraid? Would the author's career, reputation, or income be affected if the information were different?

Check 4: When was this published?

Information has an expiration date. A medical article from 2005 may have been superseded by newer research. A technology review from three years ago may describe a product that no longer exists. A historical source is valuable precisely because it is old — but a factual claim about the present world needs current evidence.

What to look for: Publication date. Whether the topic is time-sensitive. Whether the source has been updated.

Check 5: Can I verify this independently?

The strongest test of any claim: can you find the same information from a completely different source? If three unrelated sources report the same fact, it is far more likely to be true than if only one source claims it. If a claim appears in only one place and no other source confirms it, treat it with serious skepticism.

What to look for: Search the key claim in different words. Check if other reputable sources report the same thing. If the source cites a study, try to find the original study.

Build Your Scorecard Draft

Now build the tool. Rule five columns across a notebook page (use the ruler). Label them:

Scoring:

• 0 = Fails this check (no author listed, no evidence cited, obvious conflict of interest, outdated, or unverifiable)
• 1 = Partially passes (author listed but credentials unclear, some evidence but vague, mild bias, somewhat dated, partial verification)
• 2 = Passes (credentialed author, specific cited evidence, transparent motive, current, independently verified)

Maximum score: 10. Copy this template onto each of your five index cards — you will use them as portable scorecards.

Score Your First Two Sources (Practice Round)

Take Source 1 from your folder. Read it carefully. Then fill out a scorecard, writing specific observations in the Evidence and Notes columns. Do not just circle numbers — write down what you found (or did not find) for each check.

Do the same for Source 2.

Compare the two scores. Did one score significantly higher than the other? Look at which checks made the biggest difference. Are you confident in your scoring, or are there checks where you were unsure? Mark any uncertainty with a question mark — you will revisit these.

Session 2: Evaluate the Full Set (60 minutes)

The Real Test

Today you evaluate the remaining sources — six to eight of them. Work through each one methodically. Read the source. Fill out a scorecard. Write your justification for each score. Do not rush. A thorough evaluation of one source is worth more than a careless evaluation of five.

As you work, you will start noticing patterns. Certain types of sources tend to fail certain checks. Social media posts almost always fail the Evidence check. Advertisements disguised as articles often fail the Motive check. Older sources fail the Currency check even if everything else is solid. These patterns are important — write them down when you notice them.

The Hard Cases

Some sources will not fit neatly into your scoring system. Here are situations you are likely to encounter:

The source with a clear bias but solid evidence. A pharmaceutical company's press release about their own drug might cite a real, peer-reviewed study. The motive is obviously self-serving — but the evidence might be legitimate. Score the Motive check low, but score the Evidence check based on the actual evidence. Then note in your assessment: "Evidence appears solid but comes from a source with a financial interest in the outcome. Verify with independent sources."

The source with no author but useful information. Wikipedia articles rarely have a single identifiable author. Does that make them unreliable? Not necessarily — Wikipedia articles are collectively edited, sourced, and reviewed. Score the Author check low, but check whether the article has citations. Use the citations to evaluate the original sources. Wikipedia is often a good starting point for research, but it is rarely a good ending point.

The source that is technically accurate but misleading. An article might report that "crime increased 50% in Springfield this year." That sounds alarming — until you learn that crime went from 2 incidents to 3. The statistic is technically true but wildly misleading without context. This is one of the hardest things to catch. When you see percentages or statistics without raw numbers or context, mark it in your notes: "Possible misleading framing."

The source you agree with. This is the most dangerous case. When a source confirms what you already believe, your brain wants to accept it without scrutiny. This is called confirmation bias, and every human being is susceptible to it. The discipline is to evaluate sources you agree with just as rigorously as sources you disagree with. Score them honestly. If a source that supports your view gets a 4 out of 10, it is a bad source regardless of whether you like its conclusion.

Rank Your Sources

When you have scored all sources, arrange them from highest score to lowest. Write the ranked list in your notebook with the scores. Look at the ranking. Does it match your gut feeling? Are there any surprises — sources you expected to rank high that ranked low, or vice versa?

Now answer these questions in writing:

1. Which source scored highest? What specifically made it strong?
2. Which source scored lowest? What specifically made it weak?
3. Were there any sources where your gut reaction ("this seems trustworthy" or "this seems sketchy") did not match the scorecard result? What does that tell you about trusting your gut versus using a system?
4. Which of the Five Checks was hardest to evaluate? Why?

Session 3: Red Flags, Verification, and Your Final Tool (45 minutes)

Build Your Red Flags Sheet

Over the past two sessions, you have encountered dozens of specific signals — things that made you trust a source less. Now you are going to distill those into a portable reference. Take a fresh page in your notebook and title it "Red Flags: Signs of Unreliable Information."

Go through all your scored sources and your notes. Pull out every warning sign you identified. Write each one as a short, specific statement. Aim for at least ten. Here are examples of the kind of thing you might have found (do not copy these — find your own from your actual evaluations):

• No author name or bio anywhere on the page
• Uses phrases like "everyone knows" or "studies show" without naming the studies
• The URL ends in something unusual or imitates a real news site's name
• The article is trying to sell a product or service
• Extreme emotional language — designed to make you outraged, terrified, or euphoric
• No date of publication
• Claims that cannot be found in any other source
• Headline makes a dramatic claim that the article itself does not support

Your list will be different from this because it comes from your own observations. That is what makes it yours.

Verification Practice

Take the three lowest-scoring sources from your ranked list. For each one, pick the central claim — the main thing the source is asserting — and try to verify it independently. Use a library, a different website, an encyclopedia, or ask a knowledgeable adult.

For each claim, record:

• The claim, in your own words
• Where you looked to verify it
• What you found: confirmed, contradicted, or unable to verify
• If contradicted, what the more reliable source says instead

This is the verification check in action. It takes time. It takes effort. That is exactly why most people skip it — and exactly why it is powerful. The willingness to verify before believing is what separates critical thinkers from everyone else.

Finalize Your Scorecard

Look back at your scorecard template from Session 1. Based on everything you have learned, does it need revision? Maybe you want to add a sixth check. Maybe you want to adjust the scoring scale. Maybe one of the checks needs to be reworded to be clearer.

Make your revisions. Then create a clean final version — either rewrite it neatly on a fresh index card, or type it up and print it. This is your tool. You built it, you tested it on real sources, and you refined it based on experience. It works because you made it work.

Final Reflection

In your notebook, answer these questions:

1. Before this activity, how did you decide whether to believe something you read or saw online? Be specific.
2. How has your process changed? What do you do now that you did not do before?
3. What was the most surprising thing you discovered about a source during your evaluations?
4. In what situation in your real life will this skill matter most? (Think about decisions you actually make — not hypothetical ones.)
5. Is it possible for a source to score high on your scorecard and still be wrong? Why or why not?

That last question is important. Your scorecard does not tell you whether information is true. It tells you whether the source is trustworthy — whether the author, evidence, motive, currency, and verifiability all check out. A reliable source can still make an honest mistake. But the odds of being misled go down dramatically when you evaluate systematically instead of guessing.

Common Failure Modes

"Every source seems unreliable now." If you are scoring everything low, you may have set your standards impossibly high. No source is perfect. A score of 7 or 8 out of 10 is strong. The scorecard is for comparison — separating the 3s from the 8s — not for finding perfection.

"I can't tell if the author is credible." Look them up. Search their name. Do they write about this topic regularly? Do other credible sources cite them? Do they work for a relevant organization? If you cannot find any information about the author at all, that itself is useful information — it means the source is less accountable.

"I don't know enough about the topic to evaluate the evidence." You do not need to be an expert on the topic. You need to check whether evidence is present, specific, and cited. Even if you do not understand a scientific study, you can check: Is the study named? Can you find it? Was it published in a peer-reviewed journal? These are structural checks, not content checks.

"My parent says the source I scored high is actually unreliable." Good — discuss it. Ask them to point to specific things your scorecard should have caught. Maybe there is a red flag you have not learned to recognize yet. Add it to your list. The scorecard improves every time you use it.

"This takes too long to do for every source." You are right — and you will not do a full scorecard for every article you ever read. But the Five Checks will become mental habits. After enough practice, you will automatically notice when an author is not listed, when evidence is missing, when the motive is suspicious. The scorecard trains the instinct. Eventually the instinct works on its own.

Extensions

• Evaluate a full news cycle. Pick a major news event and collect coverage of it from five different sources (different outlets, different political perspectives). Score all five. Compare how the same event is framed differently depending on the source's perspective. Write a one-page analysis of the differences.
• Test your friends. Print three sources — one reliable, one unreliable, one in between — and ask a friend to rank them from most to least trustworthy. Do they identify the same red flags you would? Teach them the Five Checks and see if their ranking changes.
• Track your own media diet. For one week, keep a log of every source of information you encounter (articles, videos, social media posts, conversations). At the end of the week, estimate what percentage of those sources you would score above a 6 on your scorecard. The number may be lower than you expect.
• Evaluate AI-generated content. Ask a parent to use an AI tool to generate a short article on a topic you know well. Evaluate it with your scorecard. Where does it fail? AI-generated content often scores deceptively high on surface-level checks (clear writing, plausible-sounding claims) but fails on Author, Motive, and Verification. This is the frontier of source evaluation, and you are already equipped for it.