Facebook broke the law.
The author's bias is absurd.
1) The "research" was funded by Facebook and by money stolen from smoking cessation research funding. The "researchers" lied to the Cornell and UCSF IRB's about the funding, claiming that it was funded by the Army Research Office (which doesn't even exist) and private foundation money. O(nce the claim of Army funding was made, even if it was false, it brought the "research" under the aegis of the Office of Human Subject Protection.
2) The "researchers" lied to the Cornell and UCSF IRB's about whether it was experimental or observational science. They claimed falsely that this was only analysis of pre-existing data that was then collected. This is against the law.
3) A significant number of the participants (into the tens of thousands) were underage children who cannot give consent under Federal law. In fact, even looking at the data for such children is a crime. Children get extra protection under the law because they are exceptionally vulnerable.
4) There is a vast difference between observational research for which informed consent can be waived after review and experimental research in which informed consent, opt-out, and other laws must be followed. It becomes experimental research if anything in the subject's environment, including his news feed, is changes.
5) The "permission" cited by Facebook is for "internal operations" only. Other language was added to the TOS after the fact and hardly counts since it was not there and is essentially an admission of wrong-doing on the part of FB.
6) The PNAS action editor characterized the "research" as "creepy" and "likely a very bad example of experimental social research" but published it anyway. No one ever questioned that the paper did not disclose the sources of funding and the serious conflicts of interest involved.
7) Cornell has changed its story several times --- to the point that every point it is making is, at some point, a lie --- to try to escape liability. UCSF, UCSD, and Yale are keeping quiet and hoping it blows over.
If you want to see how these things can go awry, look up the Milgram Experiment. By causing "volunteers" to faux-electrocute other "volunteers" the author sought to demonstrate the extremes to which people will follow authority. But over half the "participants" refused to cooperate and were dropped from the study. In the remainder, a number suffered serious psychological harm from the "harmless" and "valuable" experiments. Many heads rolled over this one and a lot of the regulation goes back to this and the Tuskegee Syphilis Study.
Biting the hand that feeds IT
