Nullhypotese

Nullhypotese is a term from statistics describing a default assumption about a population parameter, typically that there is no effect, no difference, or no association. In research, it provides a baseline against which evidence from data can be evaluated. The nullhypotese is contrasted with an alternative hypothesis, which expresses what the researcher aims to show if the data provide sufficient evidence.

In formal testing, the nullhypotese is denoted H0 and the competing statement is the alternative Ha. H0 is often expressed as an equality, such as μ = μ0 or p = p0, though it can take other forms depending on the test. Researchers collect data, choose a suitable statistical test, and compute a test statistic and a p-value, which measures the compatibility of the observed data with H0.

A decision rule is applied using a predetermined significance level, commonly α = 0.05. If the p-value is less than or equal to α, the nullhypotese is rejected in favor of Ha; if not, we fail to reject H0. It is important to note that rejecting H0 does not prove Ha with certainty, and failing to reject H0 does not prove H0 is true.

Common concepts related to the nullhypotese include Type I error (rejecting a true H0) and Type II error (failing to reject a false H0), as well as one-tailed and two-tailed tests. Limitations of nullhypothesis testing include misinterpretation of p-values and sensitivity to sample size. Alternative approaches, such as confidence intervals, effect sizes, and Bayesian methods, are also used to assess evidence in data. In Dutch contexts, the term nullhypotese is widely used to describe this foundational concept.