RESEARCH · METHODS · EVIDENCE
Research self-check
Answer each question, then reveal the explanation.
1. What is the difference between random sampling and random assignment?
Reveal answer
Sampling selects people from a population; assignment allocates participants to study groups.
2. Does p = .03 mean there is a 3% probability that the null hypothesis is true?
Reveal answer
No. It describes the probability, under the specified null model and assumptions, of results at least as extreme as those observed.
3. Does a non-significant finding prove there is no effect?
Reveal answer
No. Consider the effect estimate, confidence interval, design, precision, and limitations.
4. Which reporting guidance fits a randomized trial, an observational study, and a systematic review?
Reveal answer
CONSORT, STROBE, and PRISMA, respectively. Use relevant extensions when needed.
5. Can an online questionnaire be reused without checking permissions or scoring?
Reveal answer
No. Check ownership, use conditions, version, scoring, and evidence for the intended population.
6. Why distinguish paired from independent observations?
Reveal answer
Repeated or matched measurements are related. A method that assumes independent observations can give misleading uncertainty for paired or clustered data.
References: ASA statement on p-values · NIST two-sample t-test · CONSORT reporting guidance · STROBE reporting guidance · PRISMA reporting guidance