Does Online Activity Predict Offline Behavior?

Here at KDPaine & Partners we've been studying social media and visitor engagement with social media sites, because we think it might very well have something to say about the eventual actions of the social media visitors, like whether or not they recommend or buy a company's products. But the HyveUp blog has posed an interesting question: Does online activity actually predict offline behavior? The excellent point there is that maybe people behave differently online:

"Our online life is often used as a frustration outlet... Sometimes, it just feels good to be somebody else online, or to support the candidate that it is taboo to support in your small town. Do stuff you'd never do in real life. The online world resembles a chimerical projection of our social fantasies."

This is going to take a bit of study to resolve, and my hunch is that sometimes online activity does predict offline behavior, and sometimes it doesn't. Here's some data that bears on the question, YouTube views and comments compared to voting behavior:

At least in this case, online activity can and does closely correlate with offline behavior. --Bill Paarlberg

Comments

[xavierv]

US voters have proven that they were always encouraging the most liberals, and almost always voting for the most conservatives. A Youtube graph really doesn't reflect anything. Most comments are automated spams. The Cansei de ser Sexy video could have looked like a campaign tool for Obama.

Don't you guys have correlated numbers of online engagement/vote results from past elections?

[xavierv]

Hey, me again, discussion continues
http://hyveup.blogspot.com/2008/03/online-schizophrenia-invisible-walls.html

[kdpaine]

This is really the first presidential election where YouTube played a mainstream role, so no, unfortunately, we don't have any prior data.
I would also argue that most comments are NOT automated spams, and we wouldn't include them if they were.