The greater the difference between the null hypothesis and an outcome, and the larger the sample size, the more confident one can be that the outcome is meaningful.