Contributors to this collection of previously presented conference papers discuss how the Internet and attendant hardware and software accentuate limitations endemic to polling. For example, increasing cell phone usage augments already low and declining response rates and virtually eliminates the possibility of representative samples. In addition, the authors are concerned about the greater use of Internet polling because this form of asking also produces samples, and thus, information, of extremely low, or unknown, quality. Misinformation is an acknowledged product of Internet polling.
And there’s much hand-wringing about the meaning of “public opinion”, as well as uncertainty about the effects of the Internet on it, however defined.
Although Goidel, et al. can be credited with calling attention to some effects of the Internet on political polling, a more complete analysis of the defects of polling, as well as of all other forms of asking (face-to-face interview, CASI, focus group, etc.) can be found in The Problem with Survey Research.