The following figure illustrates the principle of power analysis, consisting of four key components: Power, Alpha, sample size, and effect size (d).
First, select any of these components, such as Power, and adjust the other three to see how changes affect it. This helps you understand the relationships among these elements and the concept of power analysis.
Second, Power represents the probability of correctly rejecting a null hypothesis (H0), expressed as 1-β, where β is the probability of a Type II error (falsely accepting H0). Alpha (α), or significance level, indicates the probability of incorrectly rejecting a true null hypothesis, typically set at 0.05.
Third, sample size refers to the number of observations in the study, while effect size (d) indicates the difference from the null hypothesis.
Fourth, when holding Alpha and effect size constant, increasing the sample size results in a higher Power—larger samples provide more convincing evidence that "what is wrong is indeed wrong." Conversely, if sample size and effect size are constant, a higher Alpha leads to a higher Power.