Why might significance levels be adjusted when multiple analyses are performed?

• A. To control the overall Type I error rate
• B. To inflate p-values
• C. To shorten the trial
• D. To ignore multiplicity

Answer: (A)

When you run multiple analyses, the chance of finding at least one false positive increases. This phenomenon is multiplicity. To prevent the overall probability of a type I error from rising above the pre-specified level (often 0.05) across all tests, the significance level used for each individual test is adjusted downward. This keeps the family-wise error rate under control, preserving the integrity of the conclusions. In practice, you might see methods like Bonferroni corrections or alpha-spending approaches used, especially with interim looks or multiple endpoints. The goal is to maintain a reliable overall error rate, not to inflate p-values, shorten the trial, or ignore multiplicity.