Question: We reached significance for a variation on an A/A test and we're wondering how to interpret the results?
Answer: All else equal, it's completely within the realm of possibility to reach significance for a variation on an A/A test, but it's likely an inauthentic correlation. In this case, it's best practice to run multiple A/A tests one after another to be sure. Other than that, there may have been some bias introduced into the experiment at some point as well.
Comments
0 comments
Please sign in to leave a comment.