Null Hypothesis

In statistical hypothesis testing, the hypothesis that the pattern of results found in the observed sample is not representative for (true in) the population. For example, the hypothesis that no correlation between two variables exists in the population, that no difference between two categories exists, that a factor has no effect, etc. Researchers usually intend to collect evidence that the probability that the null hypothesis is true is very low given the pattern of results found in the sample, and thus that the null hypothesis can be rejected with an acceptably low probability of error.