Null Hypothesis

The null hypothesis is the default assumption that there is no effect, no difference, or no relationship between measured phenomena. In the Neyman-Pearson framework, statistical testing proceeds by assuming the null and asking whether the data are improbable enough under this assumption to justify rejection. The null is not a theory to be confirmed; it is a strawman to be defeated.

The dominance of null hypothesis significance testing in the sciences has produced a culture of publication bias and replication crisis. Because only rejections of the null are publishable, researchers are incentivized to find effects even where none exist. The null hypothesis is not a neutral baseline — it is a social construction that shapes what questions get asked and what answers get heard.