Weak Link Between Election Polls and Online Social-Networking...

One of the most overhyped and overrated political tools is online social-networking sites.

Both the media and internet zealots love to make statements like, "The internet is changing the face of American politics", or they give political cycles monikers like, "2008 was the YouTube Election; 2010 is the Facebook Election".

Don't buy into the hype.

Statistically speaking, there is only a very weak correlation between a candidate's popularity on social-networking sites and how they do in the polls. Some current examples...

• Delaware Senate Race
  

  
  

  	Christine O'Donnell (R)	Chris Coons (D)
  # Facebook Fans	25,809	9,523
  Real-World Polls	40%	57%

  
  

  
  
  

• California Senate Race
  

  
  

  	Barbara Boxer (D)	Carly Fiorina (R)
  # Facebook Fans	39,141	18,300
  Real-World Polls	51%	46%

  
  

  
  
  

• Indiana Senate Race
  

  
  

  	Dan Coats (R)	Brad Ellsworth (D)
  # Facebook Fans	4,422	8,553
  Real-World Polls	57%	40%

  
  

  
  
  

• West Virginia Senate Race
  

  
  

  
  

  	Joe Manchin (D)	John Raese (R)
  # Facebook Fans	5,580	3,873
  Real-World Polls	49%	49%

  
  

If all you saw were these statistics, you're only conclusion would be... Nothing. You can't draw any conclusions about the correlation between social-network popularity and election polls because, strictly going by the numbers, there isn't any.

Sure, if we expand our sample we would likely find that, overall, those candidates with more Facebook Fans tend to perform better in the polls as well. But that's weak at best. The truth is that even when that correlation is present, it's more a reflection of popular sentiment than it is a cause.