In the now decade long history of Facebook we’ve had numerous scandals around the ideas of privacy and what Facebook should and should not be doing with the data they have on us. For the most part I’ve tended to side with Facebook as whilst I share everyone’s concerns use of the platform is voluntary in nature and should you highly object to what they’re doing you’re free to not use them. The fact is that any service provided to you free of charge needs to make revenue somewhere and for Facebook that comes from your data. However this doesn’t seem to stop people from being outraged at something Facebook does with almost clockwork regularity, the most recent of which was tinkering with people’s feeds to see if emotions could spread like the plague.
The results are interesting as they show that emotions can spread through social networks without the need for direct interaction, it can happen by just reading status updates. The experimenters sought to verify this by manipulating the news feeds of some 689,000 Facebook users to skew the emotional content in one direction and then saw how the user’s emotional state fared further down the line. The results confirmed their initial hypothesis showing that emotions expressed on Facebook can spread to others. Whilst it’s not going to cause a pandemic of ecstasy or sudden whirlwind of depression cases worldwide the evidence is there to suggest that your friend’s sentiment on Facebook does influence your own emotional state.
Whilst it’s always nice to get data that you can draw causal links from (like with this experiment) I do wonder why they bothered to do this when they could’ve done much more in depth analysis on a much larger subset of the data. They could have just as easily taken a much larger data set, classified it in the same way and then done the required analysis. This somewhat sneaks around the rather contentious issue of informed consent when it comes to experiments like this as there’s no indication that Facebook approached these individuals before including them in the experiment.
Indeed that’s probably the only issue I have with Facebook doing this as whilst the data they have is theirs to do with as they see fit (within the guidelines of privacy regulations) attempting to alter people’s emotional state is a little too far. The people behind the study have came out and said that the real impact wasn’t that great and it was all done in aid of making their product better something which I’m sure is of little comfort to those who object to the experiment in the first place. Whilst the argument can be made that Facebook already manipulates users feeds (since you don’t see everything that your friends post anymore) doing so for site usability/user engagement is one thing, performing experiments on them without consent is another.
If Facebook wants to continue these kinds of experiments then they should really start taking steps to make sure that its user base is aware of what might be happening to them. Whilst I’m sure people would still take issue to Facebook doing widespread analysis on user’s emotional state it would be a far cry from what they did with this experiment, one that would likely not run afoul of established experimental standards. The researchers have said they’ll take the reaction to these results under advisement which hopefully means that they might be more respectful of their user’s data in the future. However since we’re going on 10 years of Facebook doing things like this I wouldn’t hold my breath for immediate change.