It's common today to speak of partisan polarization as reflecting societal/voter polarization. Morris Fiorina argues that the two main parties, activists, and donors are polarized but voter preferences (or public opinion) still reflects a normal distribution.

Perchance now is a good time to reflect upon whether rational choice theory models that treat party polarization as the effect of voter polarization OR Fiorina's analysis is more accurate. Along with Fiorina, I hold that social media exacerbates polarization...

In fact, there's evidence that social media, though it can create polarization, also fails to map to public opinion more generally. People are less polarized away from social media than on it. So social media misrepresents or distorts what the electorate is like to a real degree.