1. Learn
    2. /
  2. Courses
    3. /
  3. A/B Testing in R

Exercise

Family-wise error rate

The likelihood of finding a Type I error increases with the number of tests being run. The significance level and number of tests being run can be used to determine the Family-wise error rate, or the likelihood of obtaining one or more false positives, or Type I errors, when running multiple tests.

Derive the Family-wise error rate for a small and large number of tests, noting the increase in likelihood of a false positive being found among the group of tests.

Instructions 1/3

undefined XP
  • 1
    • Find the family-wise error rate of running five tests with a standard 5% significance level (i.e., alpha of 0.05).
  • 2
    • Find the family-wise error rate of running 20 tests with a standard 5% significance level (i.e., alpha of 0.05).
  • 3
    • Find the family-wise error rate of running 20 tests with a 1% significance level (i.e., alpha of 0.01).