You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is for a new feature in the Data tab of experiments.
This feature design should be represented in the redesigned UI (not the old experiment design)
When the end user clicks on the Data tab, there should be a button or other UI element that runs the condition balance check (i.e. a chi-squared statistical test). If it returns a result that indicates that the probability of the condition assignment is unbalanced and not due to chance, it displays a warning message.
Warning message text should be brief, e.g. "The actual condition assignment doesn’t match the experiment design. Debugging recommended." (maybe we want to be stronger with language here)
Optional future case: history/audit log of times the test was run
This is for a new feature in the Data tab of experiments.
This feature design should be represented in the redesigned UI (not the old experiment design)
When the end user clicks on the Data tab, there should be a button or other UI element that runs the condition balance check (i.e. a chi-squared statistical test). If it returns a result that indicates that the probability of the condition assignment is unbalanced and not due to chance, it displays a warning message.
Warning message text should be brief, e.g. "The actual condition assignment doesn’t match the experiment design. Debugging recommended." (maybe we want to be stronger with language here)
Optional future case: history/audit log of times the test was run
Relevant paper:
sequential chi-squared.pdf
The text was updated successfully, but these errors were encountered: