If Facebook were a country, it would be the third largest in the world with a population of over 1 billion. I would imagine it to secretly spy on its citizens, reading their personal communication and conducting secret experimental tests. The government would change laws and regulations frequently without notice or hesitation.
A few weeks ago, a number of my fellow U of T Digital Strategy colleagues sparked a debate about the ethicalness of a recent Facebook study. Facebook recently revealed that they anonymously conducted an emotional research experiment on unknowing users in 2012. As part of this experiment, Facebook “adjusted” the news feeds of 700K users to test the psychological implications of seeing more/less positive or negative content on newsfeeds.
Having some experience in Market Research, I felt the need to share my views on this. Firstly, to question whether a person’s emotional state can be impacted by the interaction of media elements is hardly a new discovery. Ask the romantics that cried at the end of The Notepad or the history buffs that still get goose bumps listening to Winston Churchill’s speech broadcasts – media can impact your emotions.
For the most part, we can control our exposure to certain media steams. You choose to see that movie, read that paper or visit that website. Herein lies the issue with Facebook’s approach to this research. According to Facebook, users implicitly agreed to partake as part of the general T&Cs which include this verbiage “… agreed we may use the information we receive about you … for internal operations, including troubleshooting, data analysis, testing, research and service improvement …” While Facebook may have covered itself legally, most Researchers will tell you that Facebook failed in respect to some of our industry’s guiding ethical principles:
- Secure User Opt in: I’d say they fell 700,000 short here
- Honesty & Integrity: The process had no transparency. The results themselves were only released years after the tests were conducted
- Social Responsibility, Do no Harm: Harm can be defined as both physical and psychological. In this case, no benchmarks were run to determine the initial state of the participants. What if one out of the 700,000 users exposed to negative newsfeeds became more inclined to perform self-harm in the week ensuing the tests? The study also appeared to include children and young teens which raises further concerns
Putting the ethics issue aside, the research methodology has been widely debated. What’s more surprising is that the results have shown minimal impact. For a detailed analysis, I recommend reading Dr. John Grohol’s assessment.
What’s your opinion?