A threshold set for rejecting the null hypothesis in traditional statistical hypothesis testing. This defines the Type I error and is often set at 5%. Colloquially often referred to as the “p value”.