Anglais médical · p2

Interpreting statistical results and p-values in English medical articles

Ce que tu dois retenir

Chapitre 1 — Introduction to p-values

p < 0.05 = statistically significant in most medical journals.

Chapitre 2 — Reading p-values in context

Always read p-value with its confidence interval.

Chapitre 3 — Common phrases in medical articles

Common verbs: find, report, show, observe.

Chapitre 4 — Avoiding misinterpretations

p-value ≠ probability that H0 is true.

Teste-toi

1. What does a p-value of 0.03 mean?

AA. There is a 3% chance the null hypothesis is true.
BB. There is a 3% chance of observing the data if the null hypothesis is true.
CC. The effect size is 3%.
DD. The result is definitely clinically important.
Voir la réponse

Réponse : B — B. There is a 3% chance of observing the data if the null hypothesis is true.

A p-value is the probability of observing the data (or more extreme) assuming the null hypothesis is true, not the probability that the null hypothesis is true.

2. Which p-value is considered statistically significant at the 0.05 threshold?

AA. 0.06
BB. 0.50
CC. 0.049
DD. 0.10
Voir la réponse

Réponse : C — C. 0.049

0.049 is below the usual threshold of 0.05, so it is considered statistically significant.

3. A 95% confidence interval for an odds ratio is 0.8 to 1.5. What does this suggest?

AA. The result is statistically significant.
BB. The result is not statistically significant because the interval includes 1.
CC. The p-value must be <0.05.
DD. The effect size is large.
Voir la réponse

Réponse : B — B. The result is not statistically significant because the interval includes 1.

An odds ratio of 1 indicates no effect. Since the 95% CI includes 1 (0.8 to 1.5), the result is not statistically significant at the 5% level.

4. 'No significant difference was found (p=0.42)' means:

AA. There is definitely no difference.
BB. The study had enough power to detect a difference.
CC. The observed difference could be due to chance.
DD. The difference is clinically irrelevant.
Voir la réponse

Réponse : C — C. The observed difference could be due to chance.

p=0.42 means there is a 42% chance of observing such a difference if H0 is true, so the difference could easily be due to chance.

5. Which phrase would you expect in a paper reporting a significant result?

AA. 'There was no association (p=0.08)'
BB. 'The difference was not significant (p=0.12)'
CC. 'A significant reduction was observed (p=0.02)'
DD. 'The confidence interval included the null value'
Voir la réponse

Réponse : C — C. 'A significant reduction was observed (p=0.02)'

This phrase reports a significant reduction with p=0.02, below 0.05.

6. What is p-hacking?

AA. Increasing sample size to reduce p-value.
BB. Running multiple analyses and reporting only significant ones.
CC. Using a lower alpha threshold.
DD. Correcting for multiple comparisons.
Voir la réponse

Réponse : B — B. Running multiple analyses and reporting only significant ones.

P-hacking is the practice of trying multiple analyses and only reporting those that yield p<0.05.

7. The p-value is:

AA. The probability that the alternative hypothesis is true.
BB. The probability of the observed result (or more extreme) assuming H0 is true.
CC. The probability of making a type I error.
DD. The magnitude of the effect.
Voir la réponse

Réponse : B — B. The probability of the observed result (or more extreme) assuming H0 is true.

The p-value is defined as the probability of the observed result or more extreme under the null hypothesis.

8. If a study reports p=0.001, you can conclude:

AA. The effect is very large.
BB. The result is unlikely under the null hypothesis.
CC. The result is clinically important.
DD. There is no chance of error.
Voir la réponse

Réponse : B — B. The result is unlikely under the null hypothesis.

p=0.001 indicates the data are very unlikely under H0, but it says nothing about the size or clinical importance of the effect.

Retour au cours ruedusavoir.com · Apprendre, c'est avancer.