Questions: Bayesian Methods in Biostatistics

4 questions to test your understanding

Score: 0 / 4
Question 1 Multiple Choice

A frequentist 95% confidence interval for a treatment effect is [2, 8]. A Bayesian 95% credible interval with a non-informative prior is [2.1, 7.9]. What is the key interpretive difference?

AThere is no meaningful difference — both intervals contain the true value with 95% probability
BThe confidence interval means: if we repeated the study many times, 95% of computed intervals would contain the true value. The credible interval means: given the observed data and prior, there is a 95% probability the true value lies in this interval
CThe credible interval is always narrower than the confidence interval
DThe confidence interval is for the data; the credible interval is for the parameter
Question 2 Multiple Choice

A researcher uses a strong prior centered on a treatment effect of 0 (skeptical prior) in a Bayesian analysis. Critics argue this biases the results. Is this criticism valid?

AYes — any informative prior is biased and should never be used
BPartially — the prior shifts the posterior toward 0, which may be appropriate (incorporating healthy skepticism about treatment claims) or inappropriate (ignoring strong pre-existing evidence), depending on context. The choice of prior should be transparent and sensitivity-analyzed
CNo — the prior has no effect on the posterior when the sample size is large
DNo — Bayesian methods are immune to bias by construction
Question 3 True / False

Bayesian methods are particularly advantageous for clinical trial monitoring because they can compute the posterior probability that a treatment is effective at each interim analysis without requiring multiple testing corrections.

TTrue
FFalse
Question 4 Short Answer

Explain why Bayesian analysis is said to answer the question clinicians actually care about, and what makes this different from the frequentist answer.

Think about your answer, then reveal below.