Which statistical concept provides evidence that could be used in favor of the null hypothesis?

The Bayesian factor, commonly known as the Bayes factor, is the statistical concept that provides evidence regarding the support or contradiction of the null hypothesis. In Bayesian statistics, the Bayes factor quantifies the evidence provided by the data for one hypothesis compared to another, usually the null hypothesis versus an alternative hypothesis. When interpreting Bayes factors, a value significantly greater than one would suggest support for the null hypothesis, whereas values less than one indicate support for the alternative.

This approach differs fundamentally from classical hypothesis testing, which typically focuses on p-values and their relation to the rejection of the null hypothesis. The Bayes factor offers a more nuanced perspective by allowing researchers to evaluate and express evidence directly in relation to specific hypotheses, thereby facilitating a deeper understanding of how the data relate to those hypotheses.

Other statistical concepts mentioned do not serve this specific purpose. For instance, effect size relates to the magnitude of a difference or relationship and doesn't directly provide evidence for the null hypothesis. Statistical power concerns the probability of correctly rejecting the null hypothesis when it is false, which is not relevant when considering evidence in favor of the null. Descriptive statistics summarize data characteristics but do not test hypotheses and thus do not provide evidence regarding the null hypothesis.