A few days ago, two researchers from the Department of Mechanical and Aerospace Engineering at Princeton University published an article stating that “Facebook will undergo a rapid decline in the coming years, losing 80% of its peak user base between 2015 and 2017”.
Their approach was to make an analogy between the evolution of online social networks (OSNs) and epidemiological models. They modeled the amount of search queries for the terms “MySpace” and “Facebook”, data publicly available at Google’s “Google trends” site. This article was immediately cited by several media hunger for sensationalist news, without the need of revise the information since it was mentioned as a study result.
Facebook Data Scientist Mike Develin and his team posted a sarcastic –but reasoned– response. Applying the same model to the quantity of search queries for the term “Princeton” they concluded “that Princeton will have only half its current enrollment by 2018, and by 2021 it will have no students at all”. Also, to raise the bet, they showed that the search queries for the term “Air” is declining over time as well. By extrapolating this trend they concluded “that by the year 2060 there will be no air left”.
This dispute caused a lot of buzz in many social platforms generating some rather clever responses supporting Facebook standpoint:
The next drawing published by Sean Taylor is an ironic representation of an everyday extrapolation:
This chart published by Liran Nuna exhibits the correlation between Internet Explorer market share and the number of murders in the USA:
This chart published by Jon Eide shows the negative correlation between global average temperature and the –approximate– number of pirates:
The ingenious Facebook response has a great degree of humor, but it does raise the issue of the way we can influence the results of a study with our biased perception. There is an evident need –and benefit– to recognize patterns and extrapolate trends to anticipate a future state. By definition, extrapolation is an inference about the future (or about some hypothetical situation) based on known facts and observations, always outside the range of known values. This works in many cases, but the risk is to believe that trends last forever.
Another mistake they made was to publish an article without peer revision. This might be crucial in academic environments but it is also important inside a corporation ecosystem. Leveraging in available social platform, one can easily distribute and validate lessons learned, reports and other valuable information.
It is clear that knowledge is the most important asset of a university and its appropriate treatment is a key success factor. This is where a proper Knowledge Management strategy becomes important because it contains the process of capturing, developing, sharing, and effectively using organizational knowledge. The ability to cleverly manage the various types of knowledge used by both academics and non–academics in a particular decision making, is crucial for the sustainable improvement in the performance of the organization as a whole. Knowledge Management actions usually concentrate on organizational objectives such as enhanced performance, competitive advantage, innovation, the sharing of lessons learned, integration and continuous improvement.