Which of the following is a consequence of applying the Bonferroni adjustment?

• A. Increased probability of Type I error
• B. Decreased statistical power
• C. Increased effect size estimates
• D. Shorter confidence intervals

Answer: (B)

Bonferroni adjusts the threshold for declaring significance to be more stringent for each test by dividing the overall alpha by the number of tests. This keeps the likelihood of at least one false positive across all tests from getting too high, but it makes it harder to detect real effects. In other words, you gain protection against Type I errors, but you pay with reduced statistical power—the ability to reject a false null hypothesis decreases, especially when many tests are run or effects are small.

The consequences you’d notice are not an increase in effect size estimates (these come from the data, not the significance criterion), and the confidence intervals for individual effects aren’t shorter—they tend to be wider because the per-test alpha is smaller to maintain the overall error rate.