-
-
Notifications
You must be signed in to change notification settings - Fork 18
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
The Adoption of T-test Instead of Z-test Might Improve the Accuracy #445
Comments
In fact, the Z-test functionality was omitted when we were implementing the v3 backend last year, due to the exceptional and tremendous amount of load it puts onto our infrastructure at that time. We were planning to reimplement and further, integrate it with our monitoring and alerting system in order to give us the right signal to dive into deeper investigations on a potential deviation of the dataset. Keeping the issue for tracking. Thanks a lot for the precious feedback! |
We cannot know the real σ (Standard Deviation), because the standard deviation we get is from the data we get from users. Then we have to consider df (Degree of Freedom) to the test for σ-hat. T-test could make the result more accurate because it make df into consideration while z-test only exist when df tends to infinity or σ is ideally known.
Thus, the adoption of T-test is better then Z-test.
The text was updated successfully, but these errors were encountered: