Tuesday 1 July 2014

Facebook, emotions and big data.

Over the last few days all the major media outlets have run stories about how Facebook can change your emotional state. The basis for this story is a piece of research conducted in 2012. Most of the media outlets have focused on the supposed potential for emotional control the study claims facebook might have. But there is another issue. Why is Facebook doing scientific research in the first place and how did it get ethical approval for the research without our consent?





The claims made by the paper and the media are hyperbole in the extreme. The research suggests that by changing the content of your Facebook news feed they can manipulate your emotions. In fact what they have shown is they can manipulate what you post on Facebook. They have shown no link to your emotions.

Imagine you have gone to meet your friend in the pub. You have been having a good day. The sun is shining, you got a seat on the bus, you helped an elderly person with their bag. You might want to share your positivity with your friend

But when you get their you can see they look a bit down and they tell you they have just broken up with their significant other. Suddenly you might think twice about sharing your positive stories with your friend.

All Facebook have shown is these human interactions taking place online, they have not demonstrated shifting emotional states in Facebook users. These emotional shifts might be occurring. But until these experiments are correlated with some off Facebook testing this is just pure speculation.

More worrying is that Facebook are running experiments on us at all. The original paper claims Facebook users gave informed consent when they signed up for facebook. By any sensible definition of informed consent this is complete rubbish and no ethics committee should have agreed to the research going ahead. (More on this here: http://laboratorium.net/archive/2014/06/28/as_flies_to_wanton_boys )

What is clear is that Facebook has a very different definition of informed consent to any Health Practioner. They were happy to proceed with research which they thought had the potential to lower the mood of Facebook users without asking them directly if they had consent to act in this way.

The governments recent care.data scheme, involves selling medical records to private companies for them to conduct research. With big data comes huge potential to find out more about ourselves. But if our collective medical records are going to be available to whomever has the cash to purchase them, can we trust them to use them in our interest? The example from Facebook should concern us all.

No comments:

Post a Comment