June 30, 2014

The Evils of Social Media, or Confirmation Bias in the Modern Primate

One size fits all!

Looking at the news this week, the NY Times story about Facebook data being manipulated to study/alter the emotional states of subscribers caught my eye. A familiar feeling washed over me, alerting me that my own emotional state was being altered.
The good old neurotransmitter rush of confirmation bias. My brain gleefully grabbing a bit of data that fits neatly into a slot and celebrating. Feels great. But it also lets me know that this is probably something that needs a bit more critical eye.
Happy Brain! Really.
There are a number of good reasons why our primate brains are wired up to get a happy when something we have learned is confirmed. Biology, psychology, literature, advertising and politics have been gnawing on this fact for a quite a while and a lot of science has been done to try and quantify it.
That happy rush determines what bits of data from the daily flood I pluck out and peruse, how I interpret that data, and whether I choose to share it with others.

It also signals to me, as a critical thinker, that my confirmation bias is being tickled and I need to stop and really think about what this data actually means - does the data cited in the story actually mean what the story says it means, does it mean what my brain is celebrating? Or am I ignoring stuff that makes the happy go away? Having an emotional reaction to something should be a signal to STOP and really think about what is being read/heard/reported.

And that confirmation bias buzz works just as well (if not better) with things that do not make you necessarily happy. Schadenfreude, rage, indignation - all emotions that are being tweaked and should signal a need for some objective review.

So back to Facebook. I have no love for it, belong to an age that values our privacy and hates the idea that we are a product to be bought and sold. So a story that confirms my biases against the company and reinforces my attitudes about privacy is going to ring a lot of happy bells in my brain.
So a moment to look at what was actually done, what it really means, and what is different about this study than other similar studies.

First - the study was unusual in that the study subjects never gave informed consent, were not given the option of opting out of the study, and were not debriefed at the end of the study. The normal ethical standards of psychological studies were not followed. The reason was simple - they didn't have to follow them. Since Facebook is a private company (and had a very, very great interest in the study outcome) and users of Facebook have signed an agreement giving permission for their data to be used, the researchers determined that they were not required to abide by the usual standards. Those standards are generally enforced via the funding of the study itself, and Facebook was happy to fund it.

The study involved collecting data from individual timelines by the use of keywords that signaled an emotional state. Then those individual timelines were manipulated by determining the emotional states of stories coming into the timeline and attempting to influence the overall emotional state of the user by the data that they were being shown.

Manipulating news stories to create a particular emotional atmosphere is probably as old as primates gathering in groups around a fire. It has been refined into a science that is taught and practiced by corporations, institutions and politicians with great skill and subtlety.

Early Banner Ad

The difference here is simply that rather than the shotgun approach of determining what stories go on the local evening news, what is said at a press conference, what commercial is run at the Super Bowl...etc, this is targeted at YOU. Individually.
Status update.
Nobody likes the idea that they are being personally manipulated, and that their friends and acquaintances are being used to do it. We like to think that we have a greater or lesser amount of free will, and being manipulated individually violates our personal integrity.
I am a special goddam snowflake!

The study itself is pretty unimpressive as far as adding anything new to the overall understanding of human beings or emotions. It pretty much is just a proof of concept for what is going on in social media already. The biggest puzzle is why Facebook took the risk of having something that would be so distasteful and controversial be done for publication rather than just doing it quietly in-house.
So objectively, what we are seeing is advertising targeting and emotional manipulation (especially of those confirmation biases!) simply becoming more precise. Which means that we need to be aware that this sort of individual emotional "shaping" is going on and really question the overall goals of any story that triggers our happy bias juice!

Nerd Rage, antique edition

No comments: