Statistics: the two “five” criteria
Self-tutoring about the binomial distribution: the tutor mentions a couple of criteria for approximating it to the normal distribution.
The binomial, given enough trials, settles into a normal distribution. Yet, how many trials, or what other conditions, might suggest this is safe to assume for a given case?
My reading has reminded me that when np and n(1-p) are both greater than 5, one is fairly safe to assume the binomial distribution can be appropriately approximated to the normal one. Some people caution that, as p gets further from 0.5, it’s advisable to insist on a higher value than 5.
Source:
Jack of Oracle Tutoring by Jack and Diane, Campbell River, BC.
Leave a Reply
You must be logged in to post a comment.