Home / Resources / Randy Neil Biostats

High-Yield Biostatistics Vault

Randy Neil Biostats

The definitive high-yield biostatistics study vault for USMLE Step 1/2CK/3 and COMLEX Level 1/2/3. Master "The Box", diagnostic test parameters, risk ratios, and error concepts with interactive calculations and 35 complete lesson notes.

Independent educational study guide. Audio, slides, and original video content remain the intellectual property of Dr. Randy Neil.

Interactive 2×2 Matrix Formula Cheatsheet 35 Lesson Notes Official YouTube

Lesson Notes & Transcripts

2×2

Interactive Contingency Box

10+

High-Yield Board Formulas

100%

Free & Open Access

Interactive Tool

The 2×2 Contingency Matrix ("The Box")

The foundation of board exam diagnostic questions. Adjust the cell values below or choose an exam scenario preset to watch Sensitivity, Specificity, PPV, and NPV calculate in real time.

Presets:

Contingency Table Inputs

Values update instantly

Test \ Disease	Disease Positive (D+)	Disease Negative (D-)	Row Total
Test Positive (T+)	True Positive (A)	False Positive (B)	180 Total T+
Test Negative (T-)	False Negative (C)	True Negative (D)	820 Total T-
Col Total	100 Total D+	900 Total D-	1,000 Sample (N)

Sensitivity: A / (A + C)

Specificity: D / (B + D)

PPV: A / (A + B)

NPV: D / (C + D)

Sensitivity A / (A+C)

90.0%

True Positive Rate (rules OUT disease when negative)

Specificity D / (B+D)

90.0%

True Negative Rate (rules IN disease when positive)

PPV A / (A+B)

50.0%

Chance patient has disease given positive test

NPV D / (C+D)

98.8%

Chance patient is healthy given negative test

Prevalence 10.0%

Accuracy 90.0%

LR (+) 9.00

LR (-) 0.11

SnNOut (Screening)

High Sensitivity: when test is Negative, rules Out disease. Minimizes false negatives. Ideal for population screening.

SpPIn (Confirmation)

High Specificity: when test is Positive, rules In disease. Minimizes false positives. Ideal for confirmatory testing before toxic therapy.

The Prevalence Shift

Higher prevalence increases PPV and decreases NPV. Crucially, Sensitivity and Specificity do NOT change with disease prevalence!

Exam Reference

High-Yield Board Formula Vault

The exact formulas tested repeatedly across USMLE Step 1, Step 2CK, and COMLEX. Bookmark this cheatsheet for rapid review.

Risk Reduction & Ratios

Cohort Studies

Comparing disease event rates between exposed (EER) and control (CER) cohorts.

Absolute Risk Reduction (ARR)

ARR = CER - EER

Relative Risk (RR)

RR = EER / CER

Relative Risk Reduction (RRR)

RRR = 1 - RR = ARR / CER

Exam Trap: RRR often sounds dramatic (e.g. 50% drop), but always calculate ARR to evaluate true clinical utility.

NNT & NNH

Clinical Impact

Number of patients who must be treated to prevent one bad outcome or cause one harm.

Number Needed to Treat (NNT)

NNT = 1 / ARR

Number Needed to Harm (NNH)

NNH = 1 / AR

Where AR (Attributable Risk) = EER - CER

Rounding Rule: Always round NNT and NNH UP to the next whole number (e.g. 7.2 → 8 patients).

Odds Ratio (OR)

Case-Control

Odds of exposure in cases divided by odds of exposure in disease-free controls.

Odds Ratio Formula

OR = (a × d) / (b × c)

Cross-multiplication product of the 2×2 box

Interpretation

OR = 1: No association

OR > 1: Increased exposure odds in cases

Rare Disease Assumption: When disease prevalence is low (<5%), OR closely approximates Relative Risk (RR).

Gaussian Distribution

68–95–99.7 Rule

Standard deviation percentages for symmetrical, bell-shaped normal distributions.

Mean ± 1 SD 68.2% (34.1% each side)

Mean ± 2 SD 95.4% (1.96 SD for 95% CI)

Mean ± 3 SD 99.7%

Standard Error of Mean (SEM): SEM = SD / √n. Used to construct Confidence Intervals (CI = Mean ± Z × SEM).

Type I & II Errors, Power

Hypothesis Testing

Decision matrix when evaluating the null hypothesis (H₀).

Type I Error (α) — False Positive

Rejecting H₀ when H₀ is actually true ("crying wolf"). Controlled by p-value threshold (p < 0.05).

Type II Error (β) — False Negative

Failing to reject H₀ when a true effect exists ("blindness").

Statistical Power (1 - β)

Probability of detecting a true difference. Increase sample size (n) to boost power.

Key Board Takeaway: Increasing sample size (n) increases statistical power and narrows the confidence interval.

Statistical Test Selection

Decision Guide

Choose the right test based on the number and type of variables.

Student's t-test (2 Means)

Compares means of 2 continuous groups (e.g. BP in drug group vs placebo).

ANOVA (3+ Means)

Compares means across 3 or more groups (e.g. low vs medium vs high dose).

Chi-Square (χ²) / Fisher's Exact

Compares categorical proportions in a 2×2 or r×c table.

Correlation: Pearson r measures linear association between 2 continuous variables (-1 to +1). r² is coefficient of determination.

Video Vault

35 High-Yield Video Lessons

Complete episode transcriptions, worked practice problems, and rapid recall notes. Filter by category or search by topic.

Showing 35 lessons

#01 Foundations