significantieniveau

Significantieniveau, or significance level, is a threshold used in statistical hypothesis testing to decide whether to reject the null hypothesis. It is the maximum probability of making a Type I error—the incorrect rejection of a true null hypothesis—that the researcher is willing to accept for a particular test. The level is denoted by alpha (α) and is typically chosen before data collection.

In practice, researchers compare the p-value of the observed data to α. If the p-value is less than or equal to α, the result is deemed statistically significant and the null hypothesis is rejected. If the p-value exceeds α, the null is not rejected. The same α is often used for two-sided tests, where the rejection region is split across both tails of the distribution, but one-sided tests may use α entirely in one tail.

Common choices for α are 0.05, 0.01, or 0.10, depending on field conventions and the consequences of a false positive. In multiple testing, the overall significance level is often adjusted (for example, Bonferroni or Holm methods) to control the family-wise error rate.

Significantieniveau is linked to the interpretation of confidence intervals: a two-sided 100(1−α)% confidence interval corresponds to a test conducted at significance level α.

Practical considerations include that lowering α reduces the probability of false positives but also decreases statistical power, and thus pre-specifying α is important to avoid data-driven choices and p-hacking. The chosen level affects study design, including required sample size and power calculations.