Type I/II Error

Type I and II errors are related to drawing false conclusions from data. A type I error means you think a change helped when it really did not; a type II error means you miss a real improvement because the test was not sensitive enough. You can control these risks by planning sample size, choosing clear goals, and avoiding early decisions based on noisy data. Good practice balances caution with speed so you ship real wins and avoid false alarms. Example: A company mistakenly runs an A/B test too long and collects data from a small portion of users twice. Unluckily, this portion includes heavy spenders, skewing the test results to seem as though they've increased overall revenue significantly. The company implements the change based on the test, thereby committing a type I error.