Increasing the sample size $n$

Decreasing the significance level $\alpha$

Choosing a smaller effect size

Using a one-tailed test instead of two-tailed