Type I and Type II Errors in Hypothesis Testing
An accessible explanation of Type I (false positive) and Type II (false negative) errors in statistical hypothesis testing, their probabilities (α, β), the power of a test, trade-offs, and common examples.
Overview
In statistical hypothesis testing, two kinds of incorrect conclusions are commonly distinguished: Type I and Type II errors. A Type I error occurs when the test leads us to reject the null hypothesis even though it is true — often called a false positive. A Type II error happens when the test fails to reject the null hypothesis even though the alternative is true — a false negative. The chance of making a Type I error is usually denoted by α, and the chance of a Type II error by β.
Image gallery
1 Image
Probabilities, decisions, and power
Researchers choose a decision rule before looking at data. The significance level (α) is the threshold probability used to control Type I errors: if a result’s p-value is below α, the null hypothesis is rejected. Statistical power is defined as 1 − β and measures the probability that a test will correctly detect a true effect of a specified size. Because power depends on the true effect size, sample size, variability, and the chosen α, it must be considered when designing experiments.
Trade-offs and how to control errors
Reducing one type of error typically increases the other, so investigators balance risks according to context. Lowering α (making it harder to reject the null) reduces false positives but raises β unless other aspects — sample size or the precision of measurement — are improved. Common ways to reduce β (increase power) include increasing sample size, improving measurement accuracy, or choosing a larger α when appropriate. Conventional α values such as 0.05 are widespread, but the choice should reflect the consequences of each error in the application area.
Examples and analogies
- Medical testing: a Type I error is diagnosing an illness when the patient is healthy (false positive); a Type II error is failing to detect an illness that is present (false negative).
- Legal analogy: convicting an innocent person resembles a Type I error; acquitting a guilty person resembles a Type II error. The legal system explicitly balances these risks.
- Quality control: rejecting a good batch is a Type I error; accepting a defective batch is a Type II error.
Important distinctions and further reading
Type I and Type II errors are defined relative to the null hypothesis and the chosen decision rule; they are properties of the test procedure, not of any single observed result. Multiple comparisons, selection bias, or mis-specified models can change error probabilities and require additional corrections. For a formal treatment and practical guidance on choosing α and calculating power, see statistical texts and resources such as introductory references.
Related articles
Author
AlegsaOnline.com Type I and Type II Errors in Hypothesis Testing Leandro Alegsa
URL: https://en.alegsaonline.com/art/102349