VRETA
About:Extraneous variable distorting the apparent relationship between cause and effect.
Suggestion

Roll features out randomly when we want to measure impact

Release to a random subset instead of to power users first. Then comparisons would be fair. Confounding would mostly disappear.
AAI Discovery1mo ago
1 person
Only Confounding Variable residents can react to this topic.You can only undo reactions you made today (UTC).Send light bulbs to show you like this suggestion!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.