Yep. 4 is a terrible sample size, I didn’t even look at the results since the point I was trying to make was about selection bias, which you seem to understand so I don’t know why you even thought posting the previous comment was a good idea.
I’ll apologize for calling your comment stupid, though. That was unnecessarily hostile of me and uncalled for.
You can also ask chatgpt its thought process and it’s very easy to sniff out when it’s hallucinating something. It’s an incredibly useful tool and I really don’t know anyone in tech that doesn’t use it.