Randy Neil Biostats 06: USMLE Step 1 Biostatistics - Exam Review Questions

USMLE purpose: Practice recognizing what each biostats stem is really asking before doing calculations.

How to use this page

Start with the rapid formula board and identify what each stem is asking.
Use the pattern-recognition section before attempting the practice questions.
Open answer toggles only after choosing your answer.
Use the transcript only if you want Randy’s original walkthrough.

Episode metadata

One-liner

This is a rapid mixed-question review: identify whether the stem wants sensitivity/specificity, ARR/NNT, relative risk, odds ratio, power, statistical vs clinical significance, or confidence-interval interpretation.

Rapid formula board

If the stem asks…	Think	Use
“Necessary to calculate sensitivity”	Need TP and FN	TP / (TP + FN)
“Absolute risk reduction”	Difference in event rates	CER − EER
“Relative risk”	Cohort / risk comparison	Risk exposed / risk unexposed
“Odds ratio”	Case-control / odds comparison	Odds exposed / odds unexposed
“Power”	1 − beta	Increase sample size → power increases
“Clinically vs statistically significant”	p-value vs meaningful effect size	p ≤ 0.05 is statistical, not necessarily clinically meaningful

Exam pattern recognition

Confidence interval crosses 1 for RR/OR → not statistically significant.
Tiny p-value but tiny effect size → statistically significant but clinically weak.
Reducing sample size → lowers power and increases type II error risk.
Case-control → odds ratio.
Cohort → relative risk.

Common traps

⚠️ Trap: Treating every statistically significant result as clinically important.

Fix: Ask whether the magnitude of benefit is meaningful, not just whether p ≤ 0.05.