Randy Neil Biostats 18: USMLE Step 1 - Viewer Requested Biostats Questions #5

Episode Notes

USMLE purpose: Use standard deviation rules and distinguish cohort from case-control designs.

How to use this page

For normal distributions, sketch the bell curve.
Memorize 68–95–99.7.
For study design, ask whether anyone had the outcome at the beginning.
Use the toggles before opening the explanations.

Episode metadata

Field	Details
Episode	Randy Neil Biostats 18
Topic	Viewer requested questions #5
Runtime	9 min
Published	2021-03-14
Source	Open YouTube video

One-liner

Normal distribution questions usually test 68–95–99.7 , while study-design questions usually test whether you can separate cohort → relative risk from case-control → odds ratio .

Normal distribution rule

Range	Percent included	Shortcut
Mean ± 1 SD	68%	34% on each side of mean
Mean ± 2 SD	95%	2.5% in each tail
Mean ± 3 SD	99.7%	0.15% in each tail

Study design recognition

Design	Starts with	Direction	Measure
Case-control	Disease vs no disease	Looks backward for exposure	Odds ratio
Cohort	Exposure vs no exposure	Follows forward for outcome	Relative risk
Cross-sectional	One point in time	Snapshot	Prevalence

Anki-style rapid recall

What percent is within 1 SD of the mean?

68%.

What percent is within 2 SD?

95%.

What measure is associated with cohort studies?

Relative risk.

Practice Questions

Question 1

A reference range is 0.08–0.12 and represents 95% around a mean of 0.10. What is the 99.7% range?

A) 0.07–0.13
B) 0.08–0.12
C) 0.09–0.11
D) 0.06–0.14

Reveal answer & explanation

Answer: A) 0.07–0.13

95% is ±2 SD. From 0.10 to 0.12 is 0.02, so 1 SD = 0.01. 99.7% is ±3 SD → 0.07–0.13.

Question 2

A study enrolls students based on caffeine intake and follows them for future MCAT scores. What design is this?

A) Case-control
B) Prospective cohort
C) Cross-sectional
D) Meta-analysis