Lesson 5 of 6

Common Misunderstandings

Every thinker encounters these traps. Learn to recognise the most common reasoning errors before they trip you up in real conversations, reports, and decisions.

Introduction

The traps are everywhere.

Even people who understand arguments well make reasoning errors under pressure. The good news: most common mistakes follow predictable patterns. Once you know the pattern, you can spot it instantly.

This lesson covers five of the most frequent and consequential misunderstandings in everyday reasoning. Each one has a real cost: bad decisions, wasted effort, and unfair conclusions.

The Five Traps

Patterns to recognise

Click each trap to learn it and see a real example.

🌫️ Evidence is not certainty Overclaiming

People often treat strong evidence as if it were proof. But even very strong inductive evidence leaves room for doubt. Certainty requires a sound deductive argument, and those are rare outside mathematics and formal logic.

"Studies show this diet works. It is proven. There is no reason to doubt it."

Studies show probability, not proof. The word "proven" is almost always an overclaim when applied to empirical findings.

Fix: Replace "proven" and "certain" with "strongly supported by evidence" or "highly probable." It is not weaker; it is more accurate.
💬 Opinion is not argument Missing reasons

Stating a conclusion forcefully is not the same as arguing for it. An argument requires at least one premise: a reason to believe the conclusion. Without reasons, you are simply asserting.

"This policy is obviously wrong. Anyone who supports it is not thinking clearly."

There is no argument here. There is a conclusion ("policy is wrong") and a dismissal of opponents, but no reasons are given.

Fix: Ask "What are the reasons for this claim?" If none are offered, the statement is an assertion, not an argument. It can be noted but not evaluated as reasoning.
📊 Correlation is not causation Causal fallacy

Two things happening together does not mean one causes the other. This is one of the most frequently misused reasoning patterns in journalism, business, and policy.

"Countries with more chocolate consumption have more Nobel Prize winners. Therefore chocolate makes you smarter."

Both variables may correlate with a third factor (wealth, education, research funding). Correlation can suggest a hypothesis worth testing, but it does not establish causation.

Fix: Ask what alternative explanations exist for the correlation. Look for controlled experiments or causal mechanisms before accepting a causal claim.
🥊 Ad hominem Personal attack

Attacking the person making the argument instead of the argument itself. Even if the criticism of the person is valid, it does not tell us whether the argument is good or bad.

"You cannot trust her view on climate policy. She works for an energy company."

The speaker's affiliation may be relevant context to note, but it does not tell you whether her actual argument is sound. The argument must be evaluated on its own merits.

Fix: Separate the person from the argument. Ask: is the reasoning valid? Are the premises true? A bad person can make a good argument, and vice versa.
🌊 Appeal to popularity Fallacy

The idea that something is true or good because many people believe it. Popularity is a social fact, not a logical one. Majorities have been wrong throughout history.

"Everyone knows this treatment works. Millions of people use it."

Widespread use tells us about social adoption, not efficacy. Before controlled trials, millions of people used treatments that were later shown to be ineffective or harmful.

Fix: Ask what evidence supports the claim beyond its popularity. Popularity can be a reason to investigate something further, but it is not evidence that it works.
Quick Checks

Test your understanding

Answer each question correctly to unlock the next one.

Q1. A study finds a strong correlation between two variables. A journalist writes: "Study proves X causes Y." What is wrong?
A Studies cannot find correlations.
B Correlation does not establish causation, and "proves" overclaims what a single study can show.
C The study must be peer-reviewed before any claim can be made.
D Nothing, if the study had a large sample size.
Q2. Someone says: "That politician is wrong about tax policy because he was convicted of fraud years ago." What error is this?
A Correlation is not causation.
B Appeal to popularity.
C Ad hominem: attacking the person instead of the argument.
D Overclaiming evidence as certainty.
Q3. What is the difference between an assertion and an argument?
A Assertions are always false; arguments are always true.
B Assertions are spoken; arguments are written.
C Assertions use strong language; arguments use hedged language.
D An assertion states a conclusion without giving reasons; an argument provides at least one premise as a reason for the conclusion.
Q4. "Everyone agrees this is the best approach. We should do it." Which fallacy is this?
A Ad hominem.
B Appeal to popularity.
C Correlation is not causation.
D Overclaiming evidence.
Q5. Why is it important to distinguish between evidence and certainty?
A Because certainty is always achievable with enough data.
B Because evidence should never be used in arguments.
C Because overclaiming certainty leads to overconfidence, resistance to revision, and bad decisions. Accurate calibration requires honest language about what evidence actually supports.
D Because certainty is a legal concept, not a logical one.
Mini-Game

Myth-Buster

Myth-Buster

You will be shown a flawed claim. Choose the best first check: what is the most important question to ask? Score 3 or more out of 5 to pass.

Progress: 1 / 5    Score: 0

Practice Round

Five more questions

Question 1 of 5
A manager says: "I do not need to address the criticism of my plan. The person raising it has been wrong before." What should you point out?
A Past mistakes are relevant context for evaluating future claims.
B The manager should address the criticism privately, not publicly.
C This is an ad hominem response. The criticism must be addressed on its merits, regardless of who raised it.
D Only subject-matter experts should be allowed to raise criticisms.
Question 2 of 5
A pharmaceutical company states: "Our drug has been used by 50 million patients. It is safe." What is the best critical response?
A Agree, since 50 million is a large sample.
B Disagree, because pharmaceutical companies are not trustworthy.
C Ask how many patients reported side effects.
D Note that widespread use (popularity) is not evidence of safety. Ask for controlled clinical trial data with safety monitoring.
Question 3 of 5
A data analyst says: "Sales went up after we changed the logo. The new logo caused the increase." What error might this contain?
A Ad hominem against the design team.
B Confusing correlation (or temporal sequence) with causation. Many other factors could explain the sales increase.
C Overclaiming certainty from a single data point.
D Appeal to popularity.
Question 4 of 5
When is it appropriate to use the word "proven" in an everyday argument about an empirical claim?
A Whenever the evidence is strong and from credible sources.
B Only when multiple studies agree.
C Rarely, if ever. "Proven" implies certainty that inductive evidence almost never provides. "Strongly supported" or "well-established" are more accurate.
D Only in mathematics or formal logic.
Question 5 of 5
A colleague dismisses your report by saying: "You just wrote this to make yourself look good." How should you respond?
A Explain your motives in detail.
B Note that this is an ad hominem: it addresses your motives, not the content or reasoning of the report. Invite them to identify specific flaws in the argument or evidence.
C Acknowledge that motives always affect conclusions.
D Withdraw the report to avoid further conflict.

Reflection

Think it through

Recall a recent argument you encountered, in a meeting, an article, or a conversation, where one of these five traps was present. Which trap was it? Did anyone point it out? What would have happened if they had?

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

PreviousLesson 4: Evaluating Arguments