Facebook apologized to researchers this week for data released years ago but only recently outed as inaccurately representing how U.S. users interact with posts and links. Reaching out via email and on a conference call with 47 people, the social media giant attempted to mitigate the harm caused by academics and analysts who have already spent about two years studying what they now say, and Facebook seems to agree, is flawed data about how misinformation spreads on its platform. The problem was identified as Facebook having underreported by about half the number of U.S. users and their data.

The flaw in the data was initially caught by Fabio Giglietto, an associate professor and social media researcher at Italy’s University of Urbino, according to The New York Times. Giglietto said he identified the discrepancy when comparing information Facebook released about its top posts in August to the data provided exclusively to researchers some two years ago.

Facebook spokeswoman Mavis Jones said the issue was “caused by a technical error, which we proactively told impacted partners about and are working swiftly to resolve.” The company explained that it is updating its results to correct the problem, which will take weeks due to the volume of data involved.

Facebook told call participants that only about 30 percent of those working with the material were relying on U.S. data for their studies, per NYT, but now the community is fearful that there may be other flaws. NYT speculates “Facebook was either negligent or, worse, actively undermining the research,” reporting complaints of “months of work” now lost, putting some doctoral degrees at risk.

This is the second time in a matter of weeks that Facebook discrepancies have been reported. On August 31, Politico reported that tens of thousands of Facebook posts involving the January 6 Capitol Hill riots disappeared from analytics platform CrowdTangle, which is owned by Facebook.

Facebook’s reputation took another ding earlier this week when The Wall Street Journal reported that the company, while professing to allow “more than three billion users to speak on equal footing with the elites of politics, culture and journalism,” in reality applies a double standard that exempts the powerful and famous from “some or all of its rules.”

The class system originated with a program “known as ‘Cross Check’ or ‘XCheck,’” that was developed as a quality-control layer covering the high-profile accounts managed by “celebrities, politicians and journalists,” WSJ writes, noting “Today, it shields millions of VIP users from the company’s normal enforcement process” including “whitelisted” users who are completely immune from Facebook enforcement.

XCheck docs were among the internal Facebook communications obtained by WSJ in preparing coverage. A person seeking federal whistleblower protection has made some of the same material available to members of Congress and the Securities and Exchange Commission.

Related:
Facebook Tried to Make Its Platform a Healthier Place. It Got Angrier Instead, The Wall Street Journal, 9/15/21