Well, ok, part of the internet. Twitter in particular:

Pew compared traditional public poll results with Twitter sentiment data around eight of the most significant political events over the last year, often finding significantly divergent results.

According to Pew, in some instances — Barack Obama’s reelection, the first presidential debate and a federal court ruling on California’s same-sex marriage ban — the reaction on Twitter was “more pro-Democratic or liberal than the balance of public opinion.” However, other events — Obama’s second inaugural speech, John Kerry’s nomination as Secretary of State and Obama’s 2012 State of the Union address — elicited a more conservative response on Twitter than they did in opinion polls.

The why’s are pretty straight forward:

Demographics: Twitter’s audience does not represent the real world. 13% of aduts use Twitter and only 3% regularly or sometimes tweet about the news. To top that off, Twitter skews left and young.

Sampling: Non-voters (under 18, non-US residents, etc) can participate in Twitter chats.

Grouping: Not everybody who tweets about politics tweets about every political event.

As the Mashable article points out: There is, of course, a potential conflict of interest here: Pew might not want Twitter infringing on its opinion-polling turf.

That said, it makes sense. Just as there’s a difference in polling the general public vs. registered voters vs. likely voters, there’s going to be a difference in those who are active on Twitter or any social media versus the public at large.

While one can use Twitter as a snapshot for the opinion of a SEGMENT of the population, it doesn’t necessarily reflect that of the full population.