if you're relying on statistics to show things in this area you're missing the complexity of the situation which a more interpretative account may be able to find meaning in. Statistics is not the be all and end all.
Sounds a lot like "that A/B test didn't give the results I wanted so lets try again with a higher p-value cutoff and fewer controls."
Sounds a lot like "that A/B test didn't give the results I wanted so lets try again with a higher p-value cutoff and fewer controls."