Much as reports of killer rats bound for the UK on board a ghost ship drifting in the Atlantic have been exaggerated to the point of being completely false, the recent highlighting of a study by two Princeton University researchers into the forthcoming demise of Facebook, appears to be driven more by a media frenzy than hard facts.
Yes, the authors of the Princeton paper, John Cannarella and Joshua A. Spechler, take a wholly scientific approach to their study but, as David Edwards, Senior Manager - Digital Marketing at Jersey-based digital agency, E-scape, points out in his recent blog post, "statistics can and will continue to be taken out of context".
Edwards' point "is that correlation, in isolation, can mean nothing and must not be taken at face value." In this case, he is making an important assertion.
The Princeton study uses search data from Google Trends and epidemiological models from the study of infectious diseases to map the rise and fall of Online Social Networks. The problem with the study is that it uses the brief life of MySpace as a historical reference point and as evidence for their study and its conclusions. In doing so, they inadvertently subject their analysis of Facebook to a historical context in which the mobile Internet barely existed and make nothing of the maturity of Facebook and its users. 47% of US users are aged over 35 and there's been 80% growth in the over 55's user base.
These are important factors because they cast doubt on the Princeton methodology. On the mobile front, the use of mobile apps to reach Facebook avoids the need to use a browser and therefore reduces the likelihood of using a Google search to find Facebook.
The age profile is important because older people are less likely to try new social networks and so once happily settled into Facebook and sharing with friends, the chances of them trying to move their entire network to a new platform is greatly reduced. This alone means that Facebook is likely to survive well beyond the 2017 predicted date of death that the Princeton scholars suggest.
In his article, David Edwards points out that beyond the attraction of relying on correlations that may in fact be unrelated, the Princeton report also falls into the trap of extrapolating without their being enough evidence. He highlights these with a variety of useful aides-memoires!
Perhaps Edwards' most important point is not so much the critique of the study's methodology but it's potential real-life effects.
"The paper by Mr. Cannarella and Mr. Spechler could have financial impact on Facebook if current shareholders take it at face value. As with all publicly traded companies, media coverage can and will continue to affect share price."
Which makes you wonder whether Princeton's real objective in releasing the study was to attract press attention rather than to add something of worth to the scientific canon.