The Bootstrap
Draw \(B\) random samples with replacement from your data. Compute the statistic on each. The distribution of \(B\) results approximates the sampling distribution — no formula required.
read more →Posts tagged: statistics
Effect Sizes and Cohen's d in Biomedical Research
A p-value answers one question: how surprised should I be if the null hypothesis were true? It says nothing about whether the effect matters clinically. A trial of 50,000 patients can return p < 0.0001 for a drug that lowers blood pressure by 0.3 mmHg — a difference invisible to any clinician ...
read more →