> There's always going to be some sort of sampling bias I don't agree, I think we'll get pretty good at this. I agree there currently is pretty much always a

> There's always going to be some sort of sampling bias I don't agree, I think we'll get pretty good at this. I agree there currently is pretty much always a sampling bias, though. Maybe the exception is when the study samples a particular group already, rather than the population at large.