Preparation library

Statistics for quantitative interviews

Understand variance, uncertainty, inference and validation with fresh practice.

Reviewed 8 September 2026 · Original QuantPrep guide

An estimate is not its uncertainty

A mean and a standard error answer different questions. Identify the population quantity, sampling process, dependence and units before choosing a formula.

Worked example: paired observations

For 25 paired differences with mean 2 and sample standard deviation 5, the standard error is 5/√25 = 1. The t statistic against zero is 2. Use the variation of the differences rather than treating the two groups as independent.

This does not prove a trading signal. Inspect pairing, independence, prespecified hypotheses and how many alternatives were tried.

Study progression

  1. Mean, variance, covariance and scaling.
  2. Sampling variation and interval interpretation.
  3. Conditional expectation and total variance.
  4. Regression, serial correlation and multiple testing.

Common traps

Do not confuse standard deviation with standard error. Do not interpret a confidence interval as posterior probability without a Bayesian model. Do not annualize performance blindly under autocorrelation. State the model behind every shortcut.

Two Sigma’s guidance identifies statistics and open-ended analysis as relevant to QR. Rehearse the reasoning aloud as well as numerically.

Try a fresh question

This optional exercise uses the actual parameterized bank.

Keep the practice connected

Train Statistics & Inference · Follow the Quant Path · Evidence standard