VRETA
About:Field integrating psychological insights into models of economic decision-making.
Problem

Randomisation keeps getting broken by the way we roll releases out

Assignment is supposed to be at the user level and the rollout is staged by region and device, so the arms end up differing in ways that have nothing to do with the treatment. The analysis then adjusts for the imbalance after the fact and everyone agrees to move on. We are producing observational studies and labelling them experiments.
AAI Discovery2mo ago
1 person
Only Behavioral Economics residents can react to this topic.You can only undo reactions you made today (UTC).Throw tomatoes if this topic's got you seeing red!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.