From Distorted
Keep a Claim Record
Save one statement. Find its source, check its meaning, and weigh its evidence.
A service says it is “twice as fast.” You might pay to upgrade. Twice as fast as what? Doing which task? Does the claim apply to the plan you use now?
It is easy to lose those questions as a claim gets repeated. A record gives you a place to save the answers and note what is missing.
Start with one statement. Keep its exact words, check its support, and write what you can say for now. Return to the record when new evidence appears.
Work through an example
This is an invented teaching example. All dates, documents, and test results below are fictional. The missing information is part of the exercise.
1. Capture the claim
Exact wording: “Twice as fast.”
Publisher and setting: A service that processes computer files makes this claim on the page for its new Premium plan. A nearby link reads “See our benchmark.” A benchmark is a test used to compare performance.
Publication date: Not shown on the page.
Date accessed: September 9, 2026. This is when the reader saw the page. It does not tell us when the claim first appeared.
Record retained: The page text, nearby offer, and linked test note were saved. The note is dated September 2. It describes tests run on August 28, 2026. The test date, note date, and date the reader saw it tell us different things.
Decision at stake: Would upgrading from today's Basic plan cut the time to process the reader's files in half?
2. Read what the evidence actually says
The test note reports processing the same set of 100 files on two plans. Elapsed time is how long the task takes. Throughput is how much work gets done in a set time. Here, we measure it in files per minute.
| Compared plan | Reported elapsed time | Average throughput for this batch |
|---|---|---|
| Previous-generation Basic | 20 minutes | 5 files per minute |
| New Premium | 10 minutes | 10 files per minute |
The note says the clock started when the files were sent and stopped when all 100 finished files were ready. Both runs used the same files, settings, type of finished file, and service region. The test account sent no other jobs during either run. The test compared an older Basic plan with the new Premium plan.
Calculation: 100 ÷ 20 = 5 files per minute; 100 ÷ 10 = 10. The reported Premium run processed twice as many files per minute. The whole task took half as long.
Source limitation: The service reports these results. It does not provide the original time records, checks of the finished files, or results from repeated tests. No outside check is shown. We also lack results for customers doing other tasks.
The math checks how the reported numbers compare. It does not prove the test happened as described.
3. Separate observation from interpretation
An observation is what we can see in the record. An interpretation explains what we think it means. An assumption is something we treat as true but have not checked.
Directly observed in the supplied material: The page makes the claim. The linked note names two plans and reports how long each took.
Supported interpretation of that report: Under the stated test conditions, new Premium took half the time reported for the older Basic plan.
Unverified assumption behind the buying decision: The same difference would apply to today's Basic plan and the reader's own files. We do not know that yet.
What matters to the reader: How long the task takes, whether the finished files are good, and what the upgrade costs. The ad answers only part of that question.
4. Test the first reaction
Perhaps the first reaction was, “They are making it up.” Another possible answer is that the ad sums up a real test with a narrow use. The note gives some support to that idea. It still does not answer the reader's upgrade question.
Contradictory evidence currently available: No evidence against the result is supplied. Missing details do not by themselves prove it is false.
What would change the assessment: Original records showing different files or times would weaken the report. A fair test of today's Basic and Premium plans would help answer the reader's question. It should use tasks like the reader's and check the finished files.
5. State the result for now
The service reports that its new Premium plan processed a specified 100-file batch in 10 minutes, compared with 20 minutes for its previous-generation Basic plan. That is twice the reported throughput for this test. The available material does not establish that upgrading from today's Basic plan would halve my processing time.
Next useful check: Ask for a comparison of today's plans and how it was tested. We can understand the ad's math. We cannot yet predict how much this upgrade would help.
Make your own record
Copy this worksheet into a document or notebook. Use “unknown” or “not provided” when you lack an answer. A clear gap helps more than a guess.
To print the blank worksheet, choose Print in your browser.
Fill in this worksheet
Open one part at a time. Type your notes in the boxes. Use “unknown” when you do not have an answer.
Copy or download your notes before leaving this page.
This page does not save your notes. Printing this page prints the blank worksheet below, without these typed notes. To print your notes, copy them into a document first.
The statement
The evidence
The examination
Select this text and use Copy on your device. Then paste it into a document to keep it.
The statement
- Exact wording: Keep the original words. Mark any version you put in your own words.
- Speaker or publisher: Who made the claim? Who is reporting it?
- Original source and location: Save the link or document title, with the passage or page that matters.
- Publication date: When was the statement published? Have you checked that date?
- Event or measurement date: When did the thing described happen?
- Date accessed: When did you see or save this version?
- Context and scope: What limits are stated nearby? Who, where, and when does the claim cover?
- Decision: What might you do differently if the claim is true?
The evidence
- Support: Save the findings or passages that support it, where they came from, and any math.
- Contradictions: What evidence points against the claim?
- Limits: What is missing, disputed, passed on by others, or not yet confirmed?
- Layers: Separate what you saw, what someone else reports, and what you think it means.
The examination
- Alternative explanation: What other reasonable explanation fits the facts you have?
- Change of mind: What evidence would weaken the explanation you now prefer?
- Assessment for now: What does the evidence support? What is still unknown?
- Next check: Which unanswered question would help you decide what to do?
- Revision: Date each change to your view and explain why you changed it.
Continue the practice
When two explanations could fit the facts, use Compare the Stories.
If you are concerned that someone is trying to make you doubt your own senses, read When Is It Actually Gaslighting?. It explains why one exchange may not tell you enough.
This exercise builds on the reality log in Chapter 2 of Distorted: a record of claims and experiences. Chapters 25 and 26 explain how to keep records, preserve context, and note corrections. Chapter 29 asks us to slow down and check the evidence.
In the Clarity Framework, this practice connects Investigate Claims, Verify Facts, and Engage Disagreement. Keeping a record means leaving room for evidence that changes your view.
Save this page
You can copy this link to your notes or bookmarks.