Q: 15
An advertiser wants to know whether campaign strategy A had significantly different performance
than campaign strategy B in terms of additional sales. The campaigns both ran at the same time
against mutually exclusive portions of the advertiser's customer base.
What is the null hypothesis of the test design?
Options
Discussion
Option C. had exactly this on my exam. Confirms what I saw.
C
Its C, B is a classic trap since it only checks equality and Facebook's docs usually write this as A >= B for null. Had a similar question in some practice reports.
C , Facebook tends to use "A >= B" for their null, not strict equality. B is a common trap since it's standard elsewhere.
Its B, since null usually means no difference, but C is sorta a Facebook trick.
Why is B considered the null here when Facebook usually phrases it as "A >= B" in their experiments?
B Matches what I've always learned for null hypotheses in statistics. Not totally sure for Facebook's test style though.
I don't think it's B here. The classic null hypothesis is usually about no difference, but in these campaign lift tests, "Sales Lift A >= Sales Lift B" (C) is actually how the test is often set up, especially when looking for significant improvement from A over B. B seems like a trap since it only covers equality. Pretty sure C matches what Facebook's test docs describe. Agree?
A is wrong, C. Facebook's null hypothesis usually frames it as "A performs at least as well as B" so that's why it's C here, not strict equality (B). Pretty sure about this, but let me know if you see different on recent exam reports.
B
Be respectful. No spam.
