Which factors affect statistical power?

• A. Significance level, sample size, effect size, and data variability.
• B. Only the p-value chosen.
• C. The observed p-value alone.
• D. The confidence interval width.

Answer: (A)

Power is the probability of detecting a real effect when one exists. It depends on four factors: the significance level you set (alpha), the number of observations (sample size), how large the true effect is, and how much variability there is in your data. Raising alpha makes it easier to declare significance, so power goes up, but at the cost of more false positives. Increasing the sample size reduces random error, which makes it easier to spot a real effect, so power increases. A larger true effect is easier to detect, boosting power, while more variability in measurements or outcomes (noise) makes it harder to distinguish the effect from random fluctuation, lowering power. The p-value and the width of the confidence interval reflect the data after the test, not the planning capacity to detect an effect, so they aren’t the direct determinants of power.