Power (statistics)

Relavent to Hypothesis Testing.

The probability of avoiding a Type II error; The probability that we will detect a relationship and correctly reject a false $H_0$.

There are (at least) three ways to increase power:

1. Use “parametric procedures”
2. Use a one-tailed test (b/c $z_{obt}$ is likely to be beyond $z_{crit}$ of 1.645 for a one-tailed test, versus 1.96 of a two-tailed test).
3. Use a larger sample size (N >= 30)