

I do think they need to have a good sample size to get results that make sense, yes.
Doing some research on it, seems like Quinnipiac’s standard polling is conducted “in more than 20 states and cities as well as nationally” (particular locations not specified) and has typically 1,000+ respondents using random digit dialing. Meaning they just slap random numbers into the phone and see who picks up.
I can’t say I can immediately come up with a better solution right off the top of my head right now, but I can identify several problems with their approach, namely being actual size of sample (1000 respondents is 0.000003% of the population) and the way they gather their responses (when was the last time you answered the phone for an unknown caller? Now when was the last time your grandmother did?).
English is a mish-mash hodgepodge of two dozen other languages, many (most?) of which are Romantic/Latin-based.