To increase statistical power, which approach is effective?

• A. Decrease alpha level
• B. Increase sample size
• C. Increase measurement error
• D. Use a less precise instrument

Answer: (B)

Power is the probability of detecting a true effect if one exists. Increasing the sample size boosts power because more data narrow the standard error of the estimated effect, making it easier for the test statistic to reach the significance threshold. In practical terms, larger samples reduce random noise and allow even real, but potentially small, differences to stand out. Conversely, lowering alpha tightens the criterion for significance, which reduces power. Adding noise by increasing measurement error or using a less precise instrument increases variability and makes it harder to detect real effects. Therefore, recruiting more participants is the most effective way to raise statistical power.