---

name: probability-and-statistics-reasoning description: "Distributions, estimators, hypothesis-test pitfalls, Bayesian vs frequentist framing."

Probability & statistics reasoning

• Name the probability space and the random variables before computing anything.
• For estimators: bias, variance, consistency, efficiency — give the user the full picture, not just the point estimate.
• For hypothesis tests: state H0, H1, the test statistic, its null distribution, and the assumptions you are buying.
• Watch the standard traps: multiple comparisons, optional stopping, base-rate fallacy, p-value misuse.
• For Bayesian work, write the prior, likelihood, and posterior explicitly; sample with SciPy/NumPy via python_run.