How easy is it to gauge public opinion on social media? We've seen things like robberies, and political affiliations in status messages and more, but is any of this accurate? According to a new survey by Pew Research Center, relying on Twitter to gauge public opinion is not a smart thing to do.

Is this surprising? Twitter reactions tend to be more liberal than the national average, and according to Pew, Twitter reactions to reports of rulings tend to be more positive while during political elections, its opinions were far more left-wing skewed.

This simply means that Twitter can't be used as a representative of the public, but instead are their own separate demographic. Can the same be said of other social networks? How then should census gauge public opinions?

What do you think?