Nice article. Just one thing, since you are estimating a probability, a BinomialTest is more appropriate than a OneSampleTTest. This will output a confidence interval in [0,1] and be more accurate, especially at lower numbers of samples. This addresses the first part of your conclusion.
Nice article. Just one thing, since you are estimating a probability, a
BinomialTest
is more appropriate than aOneSampleTTest
. This will output a confidence interval in [0,1] and be more accurate, especially at lower numbers of samples. This addresses the first part of your conclusion.Thank you. I didn't know that and added a comment in the article.