Bonferroni Correction: The Multiple Comparisons Problem in A/B Testing

Let’s set the stage. A CRO team runs an A/B/n test with one control and four variations, tracking three success metrics across each layout. One metric on Variant C clocks in with a p-value of 0.04.