Home » Technology » Facebook » Scientific Research Proved To Be Wrong Due To Flawed Facebook Data

Scientific Research Proved To Be Wrong Due To Flawed Facebook Data

For Facebook, the targeted collaboration with social scientists should mark a new beginning after the gloomy past in dealing with data. Now it can be seen that the data set supplied, on which many scientific papers are based, was grossly wrong.

Huge data regarding voting influence was Flawed

Facebook announced last Friday that it was a rather subdued announcement, but the consequences could be far-reaching: According to the Washington Post, the company stated that a huge data set was available to a US consortium of social scientists last year had asked, grossly flawed. The problem: The data form the basis for a large number of scientific papers, the conclusions of which must of course be called into question.

This affects the collaboration with the Social Science One consortium, founded in 2018, which Facebook actually likes to cite as a showcase project for collaboration between the company and science. Facebook had provided the researchers with data that should allow an investigation of the effect of the social network on elections and the democratic process in the United States.

The problem: The company said it had excluded all users from the database “because of an error” who could not be assigned any discernible political tendencies. It turns out that this group comprises half of Facebook users in the United States. Also unpleasant: It was not the company itself, but the Italian researcher Fabio Giglietto who became aware of the inconsistencies in the data set and informed Facebook.

Technical error, real scandal

In a statement, the company simply cites a hiccup in the system: “The problem was caused by a technical error in our URL shares data set, about which we have proactively informed the affected partners and which we are working to rectify,” said Facebook – Speaker Mavis Jones.

Of course, this news is of little consolation for the scientific community concerned. Social Science One had forwarded the erroneous data set to around 110 researchers. According to Gary King, a Harvard professor and in the management team of the research group, Facebook now urgently needs to provide the correct data in order to assess the scope.

Nathaniel Persily, Stanford Professor of Law and former chairman of Social Science One, finds even more drastic words here than his colleague: “This is an absolute scandal and a fundamental breach of the promises Facebook has made to the research community. It also shows why we need government regulation to force social media companies to develop secure data exchange programs with independent researchers.”