Post #1800703

Just saw a post (which somehow I can’t link to) about the “pitfalls of A/B testing”, and it more or less proves to me that training in #datascience doesn’t include the “science” part. They were using bad jargon but basically discussing sample bias and unintentional p-hacking. All expressed as experienced wisdom when that’s just… experiment design. It’s pretty clear that the authors were rediscovering the scientific method, painfully.