Power of the One-Sample t-Test

The one-sample t-test is used when the variance is unknown and is estimated from the sample. Its power depends on the noncentral t distribution under the alternative.

Compared with the z-test, this note reflects the extra uncertainty from estimating the variance.

R

power.t.test(delta = 0.15, sd = s, n = n,
             sig.level = 0.05,
             type = "one.sample",
             alternative = "two.sided")

Related

• STA305
• Power of a Statistical Test
• Power of the One-Sample z-Test
• Sample Size Determination