I think Quinnipaic is one of the better ones. I just know as an engineer there are limitations to extrapolation of a flawed data set. I can’t imagine that pollsters have some magical tool that wipes away the bias inherent when whole demographic groups don’t participate.
In my industry (medical devices), we lived through (and still deal with) decades of clinical studies that ignored women and BIPOC people. As a result, thousands of devices have been designed and validated that either don’t work, don’t work the same, or can outright harm people from thise communities.
That’s a flaw with a lot of polls, but I read the section on this poll’s methods and they call both landline and mobile phones.
