In ‘Power of test’ class in week4 , I see Luis mentioned power of test is not affected by sample size, however I am reading a book now ‘trustworthy online controlled experiments’ and read that in AB testing, sample size is calculated by requiring power of test to be 80%-90%, I have difficulty understanding why sample size would impact power of test, could someone help me understand this?
There seems to be an error in that content. Indeed, sample size is positively related to power (i.e., a small sample has low power, while a large sample has high power).