null hypothesis

views updated

null hypothesis The principle underlying significance tests, which proposes that any differences observed between two samples, or sub-sections of one sample, are in fact due to random sampling error rather than to real differences.