You really got to stop relying on these polls, every single poll is different, that's because they just randomly call people, if they call more democrats than republicans the results will always be more unfavorable. In another poll in the state it was tied:
North Carolina voters are divided on impeaching President Donald Trump, according to a new poll from Raleigh-based Public Policy Polling.
The poll, released Wednesday, found that 48% of voters support impeaching Trump while 48% oppose. Four percent said they were not sure.
See...it all depends upon who gets called.
You should learn how a rigorous poll is conducted. They do their best to control for the factors you mention.
There are so many variables involved in polling it's just hard to believe any of them unless a poll specifically states, (and I have seen it specifically stated) this poll was conducted of equal representation of each group. One of the problems involved in randomly chosen is that whoever hires the pollster ask them to query 1000 people randomly....that is often what happens as it would cost them more if they ask the pollster to find five hundred democrats and five hundred republicans specifically as that would require more than making 1000 phone calls. So what typically happens is a variable number of things up to and including things such as taking the number of the lesser polled and averaging them out equal as equal to what the lesser group said. So in essence that means somebody someplace that answered a phone and what they replied is considered equal to as if they had called me personally and asked me...even though there could have been a chance I may have responded differently to a question or two.
Here is a site I came across awhile back that explains how polling is done and the all different variables involved if people are curious how different polls can be conducted.
https://www.electoral-vote.com/evp2016/Info/polling-faq.html