I've long been a sceptic of the value of Facebook Likes and Twitter retweets as measures of anything but the narcissism of the people clicking on the butons.
It's largely been a gut-feeling that's occasionally backed up by anecdotal evidence. Take, for example, the experience of the well-known blogger who was delighted to receive a huge number of retweets of his tweet announcing a new blogpost and slightly less delighted to realize later that those retweets outnumbered the aggregate number of page-views the post received in the following week.
So, I was very pleased to see that we now have a statistial analysis that showed that 16% of 2.7 million studied tweets followed the same pattern.
Meanwhile, over at Facebook, the fact that Mitt Romney has lost 100,000 of his 12 million Likes in the past two weeks doesn't mean that the GOP vote has collapsed in the same time. It's not about him, it's not about voting intentions, it's about Facebook users not wanting to be associated with a losing candidate.
But where something is measurable (or at least countable), there will always be interested parties eager to asign their meaning to that which they've counted. As ever, make sure you understand what's being measured and, more importantly, why.