Skip to content

Quiz: Be Your Own Scientist: Reading Health Claims

Test your understanding of Chapter 20. Choose the best answer, then click Show Answer to check.

Question 1

What is a preprint?

  1. A short summary at the top of a published paper
  2. A complete scientific paper shared publicly before it has been peer reviewed
  3. A company announcement written to attract news coverage
  4. A plan for a trial, recorded in a registry before the trial starts
Show Answer

The correct answer is B. A preprint is a full paper posted online (for example, on medRxiv) before peer review, and medRxiv itself warns that preprints shouldn't guide health decisions. Option D describes preregistration, a different idea: it records the plan, not the finished paper.

Concept Tested: Preprint Bloom's Level: Remember

Question 2

According to a 2007 article by Hackam, about how many highly cited animal studies were later confirmed in human randomized trials?

  1. Almost all of them
  2. About nine in ten
  3. About two in three
  4. About one in three
Show Answer

The correct answer is D. Hackam summarized that only about one in three highly cited animal studies translated into confirmed results in human trials. Many people assume most animal findings carry over (options A to C), which is exactly why "a study in mice found..." deserves caution.

Concept Tested: Mouse-to-Human Translation Bloom's Level: Remember

Question 3

Which of these is a structure-function claim, the kind a US dietary supplement label can make without FDA approval (along with a disclaimer)?

  1. "Supports digestive comfort"
  2. "Treats irritable bowel syndrome"
  3. "Cures anxiety"
  4. "Prevents Crohn's disease flares"
Show Answer

The correct answer is A. A structure-function claim describes support for normal body function without claiming to treat, cure, or prevent a disease. Options B, C, and D are disease claims, which only an approved drug can legally make. The label must also say the FDA hasn't evaluated the claim.

Concept Tested: Structure-Function Claim Bloom's Level: Remember

Question 4

Why does preregistering a trial make its results more trustworthy?

  1. It guarantees that the trial's main result will be statistically significant.
  2. It means the trial was paid for by a government agency instead of a company.
  3. It records the main outcome and analysis plan before the results are known, so researchers can't quietly pick whichever outcome looked best.
  4. It means peer reviewers have already approved the paper.
Show Answer

The correct answer is C. Preregistration fixes the question, main outcome, and analysis plan in advance, which separates planned tests from after-the-fact findings. It says nothing about who paid (B) and doesn't guarantee a positive result (A) or replace peer review (D).

Concept Tested: Preregistration Bloom's Level: Understand

Question 5

A headline says a supplement "significantly reduced stress (p < 0.01)." Which statement best explains what "statistically significant" tells you?

  1. The supplement had a large effect that most people would notice.
  2. A difference this big would be unlikely by chance alone if the supplement truly did nothing, but it doesn't tell you how big or important the effect is.
  3. The study has been confirmed by other independent research teams.
  4. There is less than a 1% chance that the supplement works.
Show Answer

The correct answer is B. Statistical significance means the result would be surprising if there were no real effect. The American Statistical Association's 2016 statement says a p-value does not measure the size or importance of an effect, so option A is the classic mistake. Option D misreads what a p-value means.

Concept Tested: Statistical Significance Bloom's Level: Understand

Question 6

What does peer review usually do?

  1. Independent experts check a paper's methods, reasoning, and conclusions before a journal publishes it, though they rarely see the raw data or repeat the experiment.
  2. Independent labs repeat the experiment to confirm the result before it is published.
  3. The FDA reviews the study and approves the treatment.
  4. It proves that the paper's conclusion is true.
Show Answer

The correct answer is A. Peer review is a quality check by experts, like a building inspector who can't test every wire behind the walls. It isn't replication (B), regulatory approval (C), or proof (D); peer-reviewed studies can still be wrong, as the JB-1 mouse and human studies showed.

Concept Tested: Peer Review Bloom's Level: Understand

Question 7

In the ECOSPOR III trial, C. diff came back in 12% of people who took the microbiome treatment and 40% of people who took a placebo. What is the absolute risk difference?

  1. 12%
  2. 68%
  3. 0.32
  4. 28 percentage points
Show Answer

The correct answer is D. The absolute difference is \( 40\% - 12\% = 28 \) percentage points, meaning 28 fewer people out of every 100 had a recurrence. Option C (0.32) is the relative risk, and option B (68%) is the relative risk reduction. Both are accurate, but they describe the change as a ratio, not as actual numbers of people.

Concept Tested: Absolute vs Relative Risk Bloom's Level: Apply

Question 8

Sam's cousin shares a post: "Peer-reviewed study proves this probiotic cures anxiety!" Sam traces it to the paper and finds it was done in 12 mice, with no human data. Using the Hype Meter, what is the best grade for the post's claim as stated?

  1. Strong, because the study was peer reviewed
  2. Moderate, because it shows a real effect in animals
  3. Hype, because a small mouse study is stretched into a claim that it cures anxiety in people
  4. It can't be graded without knowing who paid for the study
Show Answer

The correct answer is C. The underlying mouse result might be interesting (Preliminary at most), but the post claims a proven human cure, which the evidence doesn't support. Peer review (A) doesn't turn a mouse study into human proof, and funding (D) matters but isn't needed to see that this claim far outruns its evidence.

Concept Tested: Claim-Checking Method Bloom's Level: Apply

Question 9

A 2017 Cochrane review (Lundh and colleagues) found that studies sponsored by the company making the product more often reported favorable results. Which conclusion best fits this finding?

  1. Company funding is a conflict of interest that calls for a closer look at design, registration, and independent confirmation, but it doesn't automatically make a study wrong.
  2. All industry-funded studies are fake and should be ignored.
  3. Funding has no effect on research results, so readers can skip the funding statement.
  4. Only government agencies are allowed to fund clinical trials.
Show Answer

The correct answer is A. The review linked industry sponsorship to more favorable results and conclusions, so a conflict of interest is a reason to look harder. But most new medicines are tested in company-funded trials, and well-designed ones (like ECOSPOR III, which was randomized, blinded, and preregistered) can still be strong evidence. Option B overcorrects; option C ignores the evidence.

Concept Tested: Conflict of Interest Bloom's Level: Analyze

Question 10

A news story says a new IBD pill "puts Crohn's into remission." The company press release cites a trial in adults whose planned main outcome was a symptom score at 12 weeks, and the results aren't yet peer reviewed. What is the most important difference between the press release and the trial record?

  1. The press release is more trustworthy because it was written by the company's experts.
  2. The trial record shows a narrower result (a symptom score in adults at 12 weeks) than the headline's broad promise of "remission" for anyone with Crohn's.
  3. There is no difference; press releases and trial records always report the same thing.
  4. The trial record proves the pill doesn't work.
Show Answer

The correct answer is B. Reading the abstract or registry shows who was studied, what was measured, and for how long, and here those are much narrower than the headline. A press release is written to attract attention (so A is wrong), and a narrow result doesn't mean the pill fails (D); it means the claim must match what was actually tested.

Concept Tested: Reading an Abstract Bloom's Level: Analyze