Learn & practice

Tools for a mind
you can trust.

Start anywhere. Take one claim from your own life and work through it. You can read every lesson without an account.

Find a tool for your next step. All six practice areas remain available below.

New here? Start with the illustrated six-step EBT routine · Learn

Your exercise entries stay in this open page. They are not sent to the project or saved between visits. Download your worksheet before leaving if you want to keep it.

01 / Emotional awareness

Make space before reacting.

A challenging idea can come with anger, relief, fear, or shame. Name the feeling and the claim separately. “This scares me” and “this is false” are different statements.

Try describing what happened, the meaning you gave it, and another possible explanation. Emotion regulation includes changing how we interpret a situation; the useful approach depends on the situation. Reframing should never require pretending harm is acceptable.

A one-minute attention practice

Choose a neutral object in the room, the feeling of your feet on the floor, or your natural breath. When attention wanders, notice that and gently return. Eyes open is fine. Nothing needs to happen.

Optional attention practice, not treatment. Skip or stop if it feels uncomfortable. Meditation does not work the same way for everyone and can cause distress for some people.

1:00

Choose a comfortable focus. Begin when you want.

Research notes: emotion regulation [1] · mindfulness benefits and limits [2]

02 / Metacognition

Notice how you arrived here.

Metacognition means examining your own thinking. The aim is to ask whether your confidence matches what you can actually show.

Ask yourself what supports a belief, what you want to be true, and what evidence you would accept against it. Feeling certain can be useful information about your state of mind. It is not a substitute for checking the claim.

Separate confidence from support

How confident are you in your chosen claim? 50%

Very doubtfulVery confident

What supports that level of confidence?What specific finding would change your mind?

This is a reflection prompt, not a score of your intelligence or a calibrated probability. You can transfer these notes to the worksheet below.

Research notes: confidence and metacognition [3] · motivated reasoning [4]

03 / Evidence literacy

Leave the page. Check the source.

A professional-looking website, confident speaker, or impressive credential can make a claim feel credible. Find out what independent sources say about the source itself.

Identify the claim. Write one clear sentence. Separate it from the speaker's conclusion.
Find the origin. Follow the citation to the study, original document, recording, or data.
Read across sources. Search outside the page. Five websites repeating one press release are still one underlying source.
Check context. Who was studied? What was measured? What is the date? What are the incentives and limitations?

Try it with a real claim

Open a second tab and investigate a source you already agree with. Find one independent account and one credible criticism. If the original evidence is inaccessible, record that limitation instead of filling in the gap.

What would a useful result look like?

“Three articles repeated one small study. It measured a short-term outcome in students, not the long-term outcome claimed in the headline. My confidence in the headline is lower.” The useful result is a clearer account of the evidence, even if your final opinion stays the same.

Research notes: research on teaching lateral reading [5]

04 / Logic & uncertainty

Does the conclusion follow?

“If this explanation were true, I might feel peace. I feel peace, so the explanation must be true.” The missing question is whether other explanations could produce the same experience. Showing that a conclusion does not follow is different from proving the conclusion false.

Good reasoning also asks what the numbers mean. How common is the thing being detected? Compared with what? What else could explain the result?

A spam warning is not a verdict

In a made-up week, 1,000 emails arrive: 10 scams and 990 ordinary messages. The filter sends all 10 scams and 90 ordinary messages to spam. There are now 100 messages in that folder.

10Actual scams in spam

90Ordinary messages in spam

100All messages in spam

A message is in spam. Does that prove it is a scam?

No. In this example, only 10 of the 100 flagged messages are scams: 10%. The other 90 are ordinary emails. Catching all the scams and correctly identifying each flagged message are different achievements. Check the sender and message before deciding.

These counts are fictional teaching numbers, not measured performance of an email service.

Teaching approach: dialogue and practical problems [6]

05 / Scientific reasoning

Ask what would distinguish the explanations.

A scientific explanation earns confidence by how well it fits observations, survives attempts to find errors, and performs against alternatives. Some questions can be tested experimentally. Others use observations, historical traces, and predictions about evidence we should find.

State the claim precisely. What is being asserted about the world?
Compare explanations. What would we expect to observe under each?
Look for an informative test. Seek evidence that can distinguish them, including evidence that could weaken your preferred account.
Check the method. Consider measurement, comparison groups, uncertainty, alternative causes, and independent checks.
Update proportionately. One weak study does not overturn a strong body of evidence. A genuine error still deserves correction.

Turn a story into a question

Someone says a remedy worked because they improved after using it. What else might explain the change? Would some people have improved anyway? What comparison would help? What outcome should be measured before looking at the results?

Use the same questions for a religious healing claim, a wellness product, or a claim made by this project.

Science can assess observable consequences. Claims defined so that no possible observation could count against them need a different philosophical discussion; being untestable does not make them established knowledge.

For reading research: NIH's guide to a scientific paper [7]

06 / Dialogue & agency

Keep the freedom to choose.

Our ethical starting point is that a person should be able to consider alternatives, ask for evidence, and refuse a demand. Agreement extracted through fear, dependency, or threatened exclusion deserves scrutiny.

In a difficult conversation, start with a specific claim. Ask permission to examine it together. Restate the other person's view fairly. Say what would change your own mind. End the discussion if either person wants to stop.

Three questions you can take with you

What evidence supports this, and can I check it independently?
Would I accept the same reasoning from a competing belief?
Can I disagree or leave without losing access to help?

A way to set a boundary

“I'm willing to talk about the evidence. I'm not willing to be threatened or pressured into agreeing. I can come back to this when we can both ask questions.”

Leaving a belief community can involve practical costs. No participant is expected to disclose doubts publicly or make an immediate decision. Learning can happen privately and at your own pace.

This lesson expresses the project's values and proposed participation standards. See participation commitments.

Bring it together

Your claim worksheet.

Make your reasoning visible to yourself. This produces a structured note from your answers. It does not decide whether the claim is true.

Your reasoning, laid out.

Download before leaving. Nothing is saved to an account.