Define statistical power.

• A. The probability of making a Type I error.
• B. The probability of detecting a true effect when it exists; typically 80–90%.
• C. The probability of obtaining a biased sample.
• D. The probability of replicating the study.

Answer: (B)

Statistical power is the probability that a study will detect a true effect if there really is one. In other words, it’s the chance of correctly rejecting the null hypothesis when the alternative hypothesis is true. Practically, researchers often aim for power around 80–90%, meaning there’s a 10–20% chance of missing a real effect (a false negative).

Power depends on several factors: how big the actual effect is, how much variability there is in the data, the sample size, and the alpha level used to declare significance. Larger samples, bigger true effects, less variability, and a less stringent alpha all increase power. Conceptually, power is 1 minus the probability of a Type II error (failing to detect an effect that exists).

The other descriptions don’t fit because: the probability of making a Type I error refers to alpha (the chance of a false positive), not power; the probability of obtaining a biased sample relates to sampling bias rather than the study’s ability to detect real effects; and the probability of replicating the study concerns reproducibility, not whether the study could identify an actual effect.