Chapter 4: Key Takeaways
The 9-Step Fact-Checker's Toolkit
| Step | Question | Why It Matters |
|---|---|---|
| 1 | What is the specific claim? | Vague claims can't be evaluated. Pin it down. |
| 2 | What is the original source? | No source = folk wisdom, not science. |
| 3 | Single study or meta-analysis? | Meta-analyses > replications > single studies. |
| 4 | What was the sample? | Size, demographics, WEIRD problem. |
| 5 | Has it been replicated? | The single most important question post-replication crisis. |
| 6 | What is the effect size? | Significant ≠ large. Most psych effects are small to medium. |
| 7 | What do other experts say? | One researcher's claim ≠ scientific consensus. |
| 8 | Who benefits? | Incentives don't mean falsehood, but they warrant scrutiny. |
| 9 | Too good to be true? | Simple, dramatic, universal claims → be suspicious. |
Core Concepts
-
The toolkit is a permanent skill. Once learned, it applies to every psychology claim you encounter for the rest of your life — on social media, in books, at work, from friends, and in this book.
-
Statistical significance ≠ practical importance. A p-value below .05 means the result is unlikely due to chance. It does not tell you the effect is large, important, or worth caring about. Always check the effect size.
-
The hierarchy of evidence matters. Meta-analyses with publication bias corrections > large pre-registered replications > multiple independent replications > single large study > single small study > no empirical evidence.
-
Unfalsifiable claims are not scientific claims. If a claim can't be stated specifically enough to be tested and potentially proven wrong, it is not a scientific claim, regardless of how it's labeled.
-
"I don't know yet" is a valid and important conclusion. Not every claim resolves neatly. Some evidence is genuinely mixed, and the honest answer is uncertainty. The ability to tolerate uncertainty is a critical thinking skill.
-
Apply the toolkit to this book too. No source is exempt from scrutiny, including this one.
Evidence Ratings in This Chapter
| Claim | Rating | Summary |
|---|---|---|
| "People are left-brained or right-brained" | ❌ DEBUNKED | No fMRI evidence for hemispheric dominance in personality |
| "Your brain replaces itself every 7 years" | ❌ DEBUNKED | Most neurons are never replaced; radiocarbon dating confirms |
| "You can learn to evaluate psychology claims with a framework" | ✅ SUPPORTED | Critical thinking instruction produces moderate improvements |
| "Couples who share housework equally are 50% less likely to divorce" | ⚠️ OVERSIMPLIFIED | Real but modest correlation; specific percentage unsupported |
Key Terms Introduced
- Effect size (Cohen's d, correlation r): The magnitude of a finding, distinct from its statistical significance
- Hierarchy of evidence: The ranking of evidence types from most reliable (meta-analyses) to least (anecdotes)
- Unfalsifiable claim: A claim that cannot, in principle, be proven wrong — and therefore is not a scientific claim
- TGTBT test: The "too good to be true" test for claims that are suspiciously clean, dramatic, or universal
One Sentence to Remember
The 9-step toolkit doesn't tell you what to believe — it tells you what questions to ask, and those questions are almost always more valuable than the answers.