Could easily be type I error. That's the problem with null hypothesis testing. You can't confidently say that the intervention caused the observed change withou

Could easily be type I error. That's the problem with null hypothesis testing. You can't confidently say that the intervention caused the observed change without replication.