Lesson 3: Signals: What is Credible? | Perspective 2

Introduction

Looking credible is not the same as being credible.

A website can have a clean design, thousands of followers, and a confident tone while containing no verifiable information. An account with a blue checkmark can spread misinformation. A post shared by thousands of people can still be wrong.

The distinction that matters: Cheap signals are things anyone can manufacture. Real evidence is something anyone can check independently.

The problem is that our brains are wired to use shortcuts. We trust confident people, popular content, and polished presentation. Online, these shortcuts can be exploited at scale. Learning to spot the difference protects your thinking.

Signal Types

Cheap signals vs real evidence

Cheap Signals

Easy to manufacture

High follower counts
Polished website design
Confident or authoritative tone
Many likes or shares
Official-sounding name
Blue verification badge
Celebrity endorsement

Real Evidence

Something anyone can check

Named, accountable authors
Cited sources with links
Transparent methods
Data that can be examined
Track record of corrections
Expert consensus across sources
Peer-reviewed research

Cheap signals feel convincing because they work as social proof in offline life. If many people are in a restaurant, it is probably good. If a speaker is confident and well-dressed, they may well know their subject. But online, these signals can be gamed with technology, purchased in bulk, or simply coincidental.

Examples

The same claim, different support

Two posts make the same claim. Click to see what their evidence actually amounts to.

Cheap Signal

"5 million people follow this account and they all know: this supplement reverses ageing. The science is settled."

Assessment: Follower count says nothing about accuracy. "The science is settled" is a confidence claim with no cited evidence. No named researchers, no studies, no methodology. This is a cheap signal stack.

Real Evidence

"A 2023 meta-analysis published in Nature Ageing (DOI: 10.1038/xxx) reviewed 42 studies and found no consistent evidence that any commercially available supplement reverses cellular ageing markers."

Assessment: Named journal, year, methodology (meta-analysis), sample size (42 studies), and a verifiable identifier. You can check every element of this claim independently.

A useful question for any claim: Can I follow a trail from this claim to something I could independently verify, or am I just being asked to trust the person saying it?

Quick Checks

Test your understanding

Answer each question correctly to unlock the next one.

Q1. An account has 2 million followers and posts confidently about nutrition. Is this a credibility signal you should trust?

A Yes, follower counts reflect public trust and expertise over time.

B No. Follower counts measure popularity, not accuracy. Popular accounts can and do spread misinformation. Look for cited sources and named expertise.

C Only if the account has been active for more than two years.

D Yes, if the account uses professional language.

Q2. A website has a clean design, a professional logo, and many internal articles. Does this indicate credibility?

A Yes. Professional presentation reflects serious effort.

B Yes, as long as the articles are long and detailed.

C No. Design is cheap and easy to replicate. Credibility comes from traceable authorship, cited sources, and accuracy across claims, not visual polish.

D Only if the website has been running for over five years.

Q3. An article cites three studies with hyperlinks and names the lead researchers. What kind of signal is this?

A A cheap signal: anyone can add fake links.

B Real evidence: the claim is verifiable. You can follow the links, read the studies, and check whether the article represents them accurately.

C Neither: studies only count if they are peer-reviewed.

D A cheap signal unless the researchers are famous.

Q4. A post has been shared 80,000 times. Does this make the claim in it more likely to be true?

A Yes. Wide sharing reflects collective fact-checking by many people.

B Yes, if the shares come from verified accounts.

C No. Shares reflect emotional resonance, not accuracy. Research consistently shows false information spreads faster than true information online.

D Only if the post is more than a week old.

Q5. Which of these is the strongest credibility signal for a health claim?

A A blue verification badge on the poster's account.

B A confident, detailed, and well-formatted post.

C A recommendation from a well-known public figure.

D A cited, peer-reviewed study with named authors and accessible methodology.

Mini-Game

Spot the Signal

Each card shows a claim or cue. Decide: is this a Cheap Signal or Real Evidence? Score at least 4 out of 6 to pass.

Progress: 1 / 6 Score: 0

Practice Round

Five more questions

Apply what you have learned. Each question unlocks after the previous answer.

Question 1 of 5

A health influencer says "I've done the research and this product works." They have 500,000 followers and a polished profile. How should you respond to this claim?

A Trust it. Large followings indicate expertise built over time.

B Trust it if the influencer has a verified badge.

C Ask what "research" means here. The claim cites no studies, no methodology, and no data. Followers and polish are cheap signals. Look for verifiable evidence before accepting the claim.

D Reject it because influencers are always advertising.

Question 2 of 5

Two friends discuss a news article. One says "it must be true, the website looks really professional." The other says "I checked two of the sources they cited and they were accurate." Whose reasoning is stronger?

A The first friend: presentation quality reflects editorial standards.

B The second friend: independently verifying cited sources is real evidence checking. Design is a cheap signal anyone can copy.

C Both are equally valid forms of evaluation.

D Neither: they should find a third source instead.

Question 3 of 5

An article has no named author, no cited sources, but its claims match something you already believe. Should you trust it?

A Yes. If the claims align with what you know, they are likely accurate.

B No. Agreement with existing beliefs is not a credibility check. It may reflect confirmation bias. Without a named author or cited sources, the claim cannot be traced or verified.

C Only if the article is from a website you have visited before.

D Yes, as long as it is consistent with other articles you have read.

Question 4 of 5

A news outlet publishes a correction when one of its articles contained an error. Does this increase or decrease your trust in the outlet?

A Decreases it. Making errors is a sign of poor journalism.

B Increases it. Publishing corrections is a mark of accountability and transparency. It is a real credibility signal: the outlet prioritises accuracy over protecting its image.

C Neither. Corrections are standard practice and say nothing about quality.

D Only increases trust if the correction was made quickly.

Question 5 of 5

You see a claim supported only by the phrase "experts say." What should you do?

A Accept it. Expert consensus is a strong credibility marker.

B Accept it if the article comes from a reputable outlet.

C Ask which experts, from where, and what they actually said. Unnamed "experts" is a cheap signal. Real expert evidence names the researchers, their institutions, and where the evidence can be found.

D Reject it. Expert claims always require peer review.

Reflection

Think it through

Think of a claim you recently found convincing online. What made it convincing? Was it real evidence you could check, or was it something that just looked or felt credible? Would your answer change now?

This is just for you. Nothing is saved or submitted.