-
Notifications
You must be signed in to change notification settings - Fork 12
Open
Description
Hi there. Nice tool - I especially like the clean interface.
I ran some simulations based its sample size recommendations, and it seemed like the recommendations were very conservative. I looked up the paper and compared with your code and it looks like you've implemented equation (3) from the paper. However, the point of the paper is that equation (3) is overly conservative, and they recommend a modification, as presented in equation (8). In my simulations, equation (8) performs better, and equation (3) consistently recommends samples that are larger than necessary by a substantial factor.
Metadata
Metadata
Assignees
Labels
No labels