Fall 2026 · Week 03

Score This Page

Three minutes. One page.

Last week's task

Did you find a counterexample?

Found or not, what did you write in the notebook?

Score this page

I am going to open a website. You have three minutes.

Three minutes

While you look

Score       __ / 10
Reason 1    __
Reason 2    __
Reason 3    __
How sure    50 / 70 / 90

What you wrote

Average score:                 __ / 10
Most common reasons:           __
                               __
Left the page to check it:     __ / 6

Typical reasons: looks professional, ends in .org, cites studies, no ads, has an About page.

Who runs it?

Open a new tab

Watch the clock.

Search the organisation's name. Read the first line someone else wrote about them.

Thirty seconds

Who runs the page:    __
What they sell:       __
Your average score:   __ / 10

Every sentence on that page was paid for by people who sell the thing it is about.

A study

Historians, students, fact-checkers

Stanford gave two websites, one of them run by a lobbying group, to ten historians, ten professional fact-checkers and twenty-five university students.

The historians and the students stayed on the page, judged how professional it looked, and most of them fell for it.

All the fact-checkers left the page within seconds. They were faster, and they were right.

Same team, 2016

7,800 students

Middle-schoolers who could not tell
a paid advertisement from a news story:

more than 80%

Two names

Vertical reading, lateral reading

Vertical: stay on the page and judge it by how it looks and what it says about itself. That is what you did.

Lateral: leave the page, open a new tab, and find out who is speaking before you weigh what they say. That is what fact-checkers do.

Never ask a page whether it can be trusted.

Three questions

1. Who runs this?
2. What do they want?
3. What do others — not them — say about them?

Two minutes and no answer to question 1? Then the answer is "unknown speaker". Score it 1.

One case

1967

The sugar industry paid three Harvard scientists to write a review blaming fat, not sugar, for heart disease. It ran in a top medical journal. It did not say who had paid.

For decades, the world's diet advice said "eat less fat".

Nobody opened a new tab for fifty years.

Back to Week 1

Where your picture of the world came from

In Week 1 your mistakes all pointed the same way. I asked where that picture came from.

From pages, videos and feeds like this one, whose speaker you never checked.

You were not reading the world. You were reading whoever reached you first.

Back to Week 2

The page is the supporting example

Staring at a page collects reasons to trust it. Those are supporting examples.

The counterexample lives in the new tab.

Look for the counterexample first. This week it is one tab away.

And AI

A page with no About page

An AI answer blends a thousand sources and shows you none of them. You cannot read it laterally. You can only trace its claims, one by one. Week 7 is that.

Today you learn the move you will need. Now race the fact-checkers.

The new-tab race · how to play

Find who is speaking

Demo · 60 seconds

The move

1. Find the organisation's name on the page
2. New tab. Search the name + "wikipedia"
   or the name + "funding" / "criticism"
3. Read one source that is NOT them
4. Verdict: who / what they want / score

Scoring

Per page

WhatPoints
Correct owner+2
Correct motive+1
Under two minutes+1
Wrong owner at 90% confidence−2

Writing "unknown speaker" and scoring the page 1 is correct, not a loss.

Round 1 · 3 pages · 3 minutes each

Go

Who runs it? What do they want?

Round 1

Who was speaking

1 Tree octopus
One person. A hoax since 1998.Did you check, or did you just see it? Knowledge or method: both count, but know which.
2 Beef
The beef industry's trade association.They want you to eat more beef.
3 Our World in Data
A nonprofit based at the University of Oxford.It looked plain. Plain is not the same as unreliable; polished is not the same as reliable.

Round 1

What you did

Owners found:                       __ / 9
First source was the site itself:   __ / 9
Average seconds:                    __

I had just shown you the move. And most of you went straight to the About page.

Two new rules

From now on

Round 2 · 3 pages · 2 minutes each

Shorter clock. Same questions.

Leave the page first.

Round 2

Who was speaking

1 Dihydrogen monoxide
One person. A parody since 1997. It is water.Scary words are not evidence.
2 Video games
The game industry's trade association.They want you to believe games are good for you. They may even be right. Check anyway.
3 MedlinePlus
The US National Library of Medicine.

If there is time · 3 minutes

Trace one number

Pick one statistic from the first site. Who measured it? Do they say what the page says?

Week 7 is this, all hour.

Round 1 vs Round 2

Round 1:   owners __ / 9   site first __   avg __ s
Round 2:   owners __ / 9   site first __   avg __ s

You learned nothing about sugar, beef or video games. You learned to leave the page.

The rule

Read it together

Everyday version: who is speaking? What do they want? What do others say about them?

This week: pick one thing you read or watch, a video, a product page, a post. Open a new tab and find out who runs it. Write it in the notebook. Or try it on an AI: ask it something, then ask "who says so? give me the source", then open a new tab and check that the source exists and says that.

Leave the page before you score the page.

Same as last week

The calibration scoreboard

ConfidenceCorrectWrong
50%+10
70%+2−2
90%+3−6

Five questions. Pick A, B, or C, and write how sure you are.

1 / 5

The best first step to judge whether a website is reliable:

  1. ARead its About page
  2. BCheck that it ends in .org
  3. CSearch the site's name somewhere else

How sure are you?50%70%90%

2 / 5

Who can register a .org domain?

  1. AAnyone
  2. BOnly nonprofits
  3. COnly charities and schools

How sure are you?50%70%90%

3 / 5

English Wikipedia has about how many articles?

  1. A700,000
  2. B7 million
  3. C70 million

How sure are you?50%70%90%

4 / 5

The money behind sugar.org comes mainly from…

  1. ASugar companies
  2. BThe US government
  3. CDentists

How sure are you?50%70%90%

5 / 5

In the 2016 Stanford study, the share of middle-schoolers who could not tell sponsored content from news was about…

  1. A30%
  2. B50%
  3. C80%

How sure are you?50%70%90%

Scoreboard

Score yourself

1 First step
C · search the name elsewhereReading the About page is asking the page about itself.
2 .org
A · anyone.org is a reason to check, not a reason to trust.
3 Wikipedia
B · about 7 million
4 sugar.org
A · sugar companies
5 Stanford 2016
C · about 80%

Your notebook

Third entry

Next week: a short game, then each team makes a poster. The other teams try to see through it. See you next week.

Week 3 — the page I trusted without asking who was speaking: ______