5 minutesA scientist who says people are warming the planet
vsvs
5 minutesOne who says they are not
Two chairs, the same time each.
A television studio. Two chairs. In one sits a scientist who says people are warming the planet. In the other sits one who says they are not. Each gets the same five minutes. Before either speaks, the screen has already told the viewer something: this question has two even sides.
Airtime for chair one50%
Airtime for chair two50%
Giving each side the same time is called balance. The word just means each side gets an equal turn. Balance is fair when the people who study a question really are split. It is not fair when nearly all of them agree, because then the equal turns show a split that is not there.
5 minutesA scientist who says people are warming the planet
vsvs
5 minutesOne who says they are not
One chair each, the same time each. Before anyone speaks, what does that tell a viewer?
Nothing. It is just a debate.
Not yet. A debate has a shape, and the shape says two sides of the same size. On a question where nearly everyone agrees, the shape is a claim, and a false one.
That the people who study this are split about evenly.
Right. One chair each is a picture of a fifty-fifty question. The viewer reads the picture before a word is said. Nobody on the programme has to say the field is split; the two chairs say it for them.
That the first speaker is right.
Not yet. The set-up favours neither chair. That is the whole problem when one chair stands for nearly everyone and the other for a few.
Of every 100 published climate studies that say what is causing the warming, how many say humans are the main cause?
Pick one. Nothing is scored.
97 of 100
studies that say humans are the main cause
About 97.Ninety-seven to three in the journals, one chair each in the studio. Nobody said anything untrue. The seating plan showed a split that does not exist.
The view held by 97 in 100 scientists5 minutes
The view held by 3 in 1005 minutes
So the studio gave the view held by 3 in 100 half the time. Nobody lied. Both guests said what they believed. But the clock told the viewer the field is split down the middle, and it is not. The viewer walks away with a picture the evidence does not support.
A ten-minute segment on a settled question. Tap the voice that got MORE time than its share of the evidence.
Tap the part of the picture that answers it.
The view held by 97 in 100 scientists5 minutes
The view held by 3 in 1005 minutes
The view held by 3 in 100
Right. Five minutes for three per cent of the evidence: about sixteen times its share. Nobody lied. The split of the minutes did the exaggerating.
The view held by 97 in 100 scientists
Not yet. Five minutes for ninety-seven per cent of the evidence. It got half the time and holds nearly all the evidence, so this is the voice that got less than its share.
Should the voting age be sixteen?what people want
vsvs
Does smoking cause cancer?what is so
Two chairs suits the first kind only.
Some questions are about what people want. Whether the voting age should be sixteen is a choice. No study can settle it, so two chairs is honest. Other questions are about what is so. Whether smoking causes cancer was settled by evidence decades ago, and two chairs would show a split that ended then.
A real open question, or one the evidence has settled?
One card at a time. Tap the pile it belongs to.
Card 1 of 6
Should the voting age be sixteen?
Do vaccines cause autism?
Is the climate warming because of us?
Should the country join a trade group with its neighbours?
Does smoking cause cancer?
Is the new stadium worth the money?
Should the voting age be sixteen? → A real open question
Not yet. This is open. It asks what people want, and no study can tell a country what it should want.
Do vaccines cause autism? → The evidence settled it
Not yet. The evidence settled this. Studies of millions of children found no link, so two chairs would invent a doubt.
Is the climate warming because of us? → The evidence settled it
Not yet. The evidence settled this. Decades of measurements point one way, and the few who disagree are not half the picture.
Should the country join a trade group with its neighbours? → A real open question
Not yet. This is open. It is a choice about what the country wants, and both sides have real reasons.
Does smoking cause cancer? → The evidence settled it
Not yet. The evidence settled this decades ago. Giving the other side a chair would make a false claim about the science.
Is the new stadium worth the money? → A real open question
Not yet. This is open. Worth it to whom, and instead of what, are questions about what people want, not about facts.
About what people wanttwo chairs is honest
vsvs
About what is sothe evidence sets the shares
Right. The open ones are about what people want, and no study can settle them. The settled ones are about what is so. Two chairs suits the first kind. On the second kind, the chairs themselves make a false claim.
What the evidence says, share for the main view97%
What the viewer takes away50%
Illustrative.
Put the two numbers side by side. How far the evidence leans one way. How the airtime is split. When the two match, the viewer's picture is right. When the airtime is more even than the evidence, the viewer sees an open question where there is a settled one.
The numbers are made up. Set how settled the evidence is, and how the airtime is split.
What the evidence says, share for the main view50%
What the viewer takes away50%
Illustrative.
How settled the evidence is: An open question · Airtime split: 50 / 50
Evidence 50% one way, airtime 50 / 50. The viewer's picture matches the evidence.
What the evidence says, share for the main view50%
What the viewer takes away70%
Illustrative.
How settled the evidence is: An open question · Airtime split: 70 / 30
Evidence 50% one way, airtime 70 / 30. The viewer's picture is 20 points too one-sided.
What the evidence says, share for the main view50%
What the viewer takes away97%
Illustrative.
How settled the evidence is: An open question · Airtime split: 97 / 3
Evidence 50% one way, airtime 97 / 3. The viewer's picture is 47 points too one-sided.
What the evidence says, share for the main view75%
What the viewer takes away50%
Illustrative.
How settled the evidence is: Leaning one way · Airtime split: 50 / 50
Evidence 75% one way, airtime 50 / 50. The viewer's picture is 25 points too even.
What the evidence says, share for the main view75%
What the viewer takes away70%
Illustrative.
How settled the evidence is: Leaning one way · Airtime split: 70 / 30
Evidence 75% one way, airtime 70 / 30. The viewer's picture is 5 points too even.
What the evidence says, share for the main view75%
What the viewer takes away97%
Illustrative.
How settled the evidence is: Leaning one way · Airtime split: 97 / 3
Evidence 75% one way, airtime 97 / 3. The viewer's picture is 22 points too one-sided.
What the evidence says, share for the main view97%
What the viewer takes away50%
Illustrative.
How settled the evidence is: Settled · Airtime split: 50 / 50
Evidence 97% one way, airtime 50 / 50. The viewer's picture is 47 points too even.
What the evidence says, share for the main view97%
What the viewer takes away70%
Illustrative.
How settled the evidence is: Settled · Airtime split: 70 / 30
Evidence 97% one way, airtime 70 / 30. The viewer's picture is 27 points too even.
What the evidence says, share for the main view97%
What the viewer takes away97%
Illustrative.
How settled the evidence is: Settled · Airtime split: 97 / 3
Evidence 97% one way, airtime 97 / 3. The viewer's picture matches the evidence.
When is a fifty-fifty split honest?
Always. Fair is fair.
Not yet. Slide the evidence to settled and the airtime to fifty-fifty: the viewer's picture is 47 points off. Fair to the two guests, and false to the audience.
When the evidence is genuinely open.
Right. The bars match only when the airtime follows the evidence. Fifty-fifty on an open question is fair. Fifty-fifty on a settled one gives the three per cent half the time.
Never. One side is always right.
Not yet. On an open question, fifty-fifty is the honest picture. The mistake is using one split for every question.
Move both controls.
Airtime for the view most scientists hold90%
Airtime for the view few hold10%
In 2011 a broadcaster reviewed its own science coverage. It found it had given too much time to views on climate that very few scientists hold. What did it change?
Nothing. Equal time is the rule.
Not yet. The rule was re-read. It had always asked for a fair share, not an equal one, and on a settled question a fair share is not half.
It started giving each view time in proportion to the evidence behind it.
Right. The review called this due weight. The view held by few is still heard. It is heard at the size it is, which on a settled question is small.
It banned the views held by few.
Not yet. Nothing was banned. A view held by three in a hundred gets a share nearer three than fifty.
Lesson complete
Giving two sides equal time tells the viewer the evidence is split, even when it is not.