We are more likely to reject the null hypothesis when

the amount of variability in the population we sampled is low.