Believing with your team
We judge evidence harder when it threatens our group or our beliefs. See how identity bends reasoning and why people rarely dig in as much as we fear.
TWO QUESTIONS
Motivated reasoning
Motivated reasoning is reasoning toward the conclusion you already want. Psychologist Tom Gilovich described it as asking two different questions. For a welcome claim we ask "Can I believe this?", and almost anything clears that bar. For an unwelcome claim we ask "Must I believe this?", and almost nothing does.
Same kind of evidence, two different standards. The standard moves; the evidence doesn't.
Nasrin follows a low-carb diet
Study finds a benefit
"Low-carb diet linked to better heart health."
She reads the headline and shares it with her sister within a minute. Can I believe this? Sure.
Similar study finds a risk
"Low-carb diet linked to higher cholesterol."
She checks who funded it, how many people took part, how long it ran. Must I believe this? Not yet.
Check yourself
Kaveh searches hard for flaws in a study only when he dislikes its result. When he likes the result, he shares it without a second look. What is this?
- Healthy scepticism
- Motivated reasoning
- Lateral reading
- The illusory truth effect
Show the answer
Motivated reasoning
Right. Checking for flaws is good; checking only the results you dislike is motivated reasoning. The standard follows his wishes, not the evidence.
WHEN A BELIEF IS A BADGE
Identity makes confirmation bias stronger
Once we lean one way, we notice agreeing evidence more and forget the rest: confirmation bias (the Mental Models course covers it fully). Identity turns the volume up. When a belief marks which group you belong to, your family, your town, your team, doubting it can feel like betraying your people, so the mind works hard not to.
A claim about a rival city's football club gets believed or doubted by fans long before anyone reads the details.
A belief worn like a team shirt
Nobody checks the league table before deciding to support their team. The shirt comes first; the reasons come after. Some beliefs work the same way: they tell others which side we're on. Where the analogy breaks: a team shirt is harmless, but a belief about a medicine or a safety warning has real consequences, and the facts don't care whose shirt you're wearing.
Check yourself
People who are very good with numbers and arguments are protected from motivated reasoning.
Show the answer
False
False. Some studies suggest skilled reasoners can be even better at arguing for the side they already prefer. Skill is a tool; it works for whatever conclusion you point it at.
EVERYONE'S SAYING IT
Social proof
We copy what people like us believe and do. Often that's sensible: a busy restaurant is usually decent. Online it misleads, because likes and shares can be bought, bots can inflate the numbers, and "everyone's saying" may be one loud group. A big number tells you a post is popular, not that it's true.
A post with 50,000 likes and a post with 50 likes can make the same claim. The likes measure attention, not evidence.
Check yourself
A post claiming a common painkiller damages memory has 50,000 likes. What does the like count tell you about whether the claim is true?
- A lot: that many people can't all be wrong
- Little: popularity isn't evidence, and the numbers can be inflated
- That doctors have probably approved it
- That it's definitely false, since viral posts always are
Show the answer
Little: popularity isn't evidence, and the numbers can be inflated
Right. Likes measure how many people reacted, not how many checked. They can also be bought or pumped up by bots. The claim still needs its own evidence.
The double-standard test
- Notice your reaction
Quick yes, or a sudden urge to find flaws? Either one is a clue that you care about the answer.
- Flip the side
Imagine the same evidence, same study size, same source, pointed at the conclusion you don't like.
- Ask: would I accept it now?
If you'd tear the flipped version apart, your standard is moving with your wishes.
- Use one standard for both
Apply the checks you'd use on the unwelcome version to the welcome one too.
Check yourself
- A small online survey finds that people in Arash's home town are the friendliest in the country. He posts it with "Science agrees."
- The same kind of small online survey finds his home town is among the least friendly. He comments: "Online surveys prove nothing."
What changed between the two reactions?
- The quality of the evidence
- Only whether Arash liked the result
- The number of people surveyed
- The survey's source
Show the answer
Only whether Arash liked the result
Exactly. Same method, same weakness. He's right that small online surveys prove little, but that applies to the first one too. That's the double standard.
Check yourself
What question does the double-standard test ask?
- Would I accept this evidence if it supported the other side?
- How many people on my side agree with this?
- Is the source on my side or theirs?
- Can I find at least one study that agrees with me?
Show the answer
Would I accept this evidence if it supported the other side?
Yes. Flip the conclusion and keep everything else the same. If your verdict flips too, the standard is doing the moving.
GOOD NEWS
The backfire effect is rare
A 2010 study by Brendan Nyhan and Jason Reifler reported a backfire effect: correcting some people seemed to make them believe the false claim more. It became famous. But later, larger studies, such as Thomas Wood and Ethan Porter's in 2019, found backfire is rare. Most of the time, corrections move people at least a little toward the facts.
So correcting is worth doing. Do it gently, and don't expect a complete change of mind in one conversation.
Common mistakes
- Thinking only the other side does this. Everyone does; the test is whether you catch yourself.
- Mistaking lots of shares for lots of evidence.
- Giving up on correcting people because "it only makes them dig in". Usually it doesn't.
Check yourself
What did later, larger studies find about the "backfire effect", the idea that corrections make people believe a false claim more?
- It happens almost every time, so correcting people is pointless
- It's rare; corrections usually move people a little toward the facts
- It only happens to people with less education
- It was never studied again after 2010
Show the answer
It's rare; corrections usually move people a little toward the facts
Right. Larger studies, like Wood and Porter's in 2019, found backfire is rare. A calm correction usually helps, even if only a little.
Lesson recap
- Motivated reasoning: "Can I believe this?" for welcome claims, "Must I believe this?" for unwelcome ones.
- When a belief marks your group, doubting it feels like leaving, which makes confirmation bias stronger.
- Social proof misleads online: likes can be bought and inflated, and popular isn't proven.
- The double-standard test: would I accept this evidence if it supported the other side?
- The backfire effect is rare; gentle corrections usually move people a little toward the facts.